Running a SAFE Financing Round: A Practical Checklist

Corporate > Securities | Published 28 December 2025 - Updated 21 February 2026 | UpLaw - uplaw.us

A working checklist for issuing simple agreements for future equity, from choosing the instrument through the four economic levers, post-money dilution mechanics, conversion and phantom-preference modeling, Regulation D compliance, side letters, and clean closing mechanics.

Keywords: SAFE, post-money SAFE, valuation cap, discount rate, MFN clause, convertible note, seed financing, Regulation D, Rule 506(b), Rule 506(c), accredited investor, Form D, blue sky laws, dilution, cap table, phantom liquidation preference, pro rata rights, Section 4(a)(2), Section 409A.

Somewhere right now, a founder is signing her fourth SAFE of the quarter, feeling productive. Each one took an afternoon. No board vote, no amended charter, no lawyer billing for a week of preferred-stock drafting. The instrument is four pages long and its name promises exactly what she wants: simple. Eighteen months later, at a Series A closing, a venture fund's counsel drops a pro forma cap table on the table and the founding team learns that the "simple" paper conveyed away thirty-eight percent of the company before the priced round issued a single share. Nobody lied. Nobody breached anything. The documents did precisely what they said. The trouble is that the most consequential thing they said was compressed into that one adjective, and the founders read it as a description of the drafting when it was actually a description of the paperwork, not the consequences.

This checklist exists to keep that meeting from being a surprise. A SAFE round is not hard, but it is unforgiving of the founder who treats conversion as a problem for future-you. Run in order, the phases below turn a fundraise into a sequence of deliberate decisions: whether to use a SAFE at all, how to set the handful of terms that actually matter, how to model the dilution before it is irreversible, how to comply with the securities laws that govern every sale, and how to close without seeding a thicket of side-letter obligations. For the doctrinal deep dive with fully worked conversion math, read the companion overview, Simple Agreements for Future Equity: An Overview of How SAFEs Actually Work. For where the SAFE sits in the wider paper stack, see The Startup Legal Document Stack: What Founders Need and Why and the stage-by-stage map in The Founder's Path from Formation Through Financing.

How to use this checklist

Work the phases sequentially, but treat Phases 3 and 4 as the load-bearing ones. Everything before them is setup; everything after is execution. The single most valuable habit this checklist can instill is building one pro forma model, before the first SAFE is signed, that answers a single question: if I raise the full amount I am planning, on the terms I am offering, what percentage of my company will I own after it all converts at the priced round? Founders who can answer that question in a number rarely get ambushed. Founders who answer it in dollars raised almost always do.

A note on vocabulary. A SAFE is not a loan and not stock. It is a contract giving the investor the right to receive stock later, when a triggering event occurs, at a price set by formulas fixed today. Until that trigger, the investor holds a piece of paper with no maturity date, no interest, no vote, and no liquidation claim of its own. That structural fact drives nearly everything that follows.

Phase 1 — Decide whether a SAFE is the right instrument

• [ ] Confirm the SAFE serves a deliberate financing strategy rather than substituting for one.
• [ ] Assess investor sophistication and preference; some angel and regional investor pools still strongly prefer convertible notes, and a few will not do SAFEs at all.
• [ ] Compare the four real options side by side: convertible note (debt, with interest and maturity), SAFE (no interest, no maturity), Series Seed convertible preferred (a small priced round), and, rarely, common stock.
• [ ] Weigh transaction cost against the amount raised; legal and filing fees can swallow a small round, which pushes toward the simplest instrument.
• [ ] Check whether any grant, accelerator, or non-dilutive funding program forbids debt-like instruments or dictates the equity structure.
• [ ] Confirm the company is a corporation, not an LLC; the standard SAFE forms assume a C-corporation with authorized preferred stock (or the room to authorize it).

Why the question comes first. The reflex to reach for a SAFE without asking whether it fits is itself the first mistake. A SAFE is company-friendly by design: it strips out the interest, the maturity date, and the security interest that protect an investor who lends through a convertible note. Whether that stripping-out is a feature or a defect depends entirely on which side of the table you occupy. Practical Law's comparison of seed instruments captures the trade cleanly: SAFEs offer "no maturity or interest," which founders prize for the peace of mind, while some investors "prefer investing in early-stage companies through convertible notes" precisely because a SAFE "has no maturity date and offers no interest." A note gives the investor leverage; a SAFE gives the founder calm. Neither is right in the abstract.

Cost is the quiet tiebreaker at the smallest end. A SAFE or a note can be documented for a fraction of a priced round, which requires longer transaction documents and an amended-and-restated certificate of incorporation, incurring state filing fees and counsel time. When the round is $250,000, spending $30,000 to paper a priced Series Seed is malpractice against the company's bank account. When the round is $3 million from an institutional lead who wants a board seat and protective provisions, the priced round's cost buys certainty that the SAFEs cannot. The decision is a function of round size, investor mix, and how much bespoke structure the parties actually need.

There is a diagnostic worth internalizing: if you find yourself wanting three custom side letters, the honest answer is that you want a priced round. The SAFE's whole value proposition is that it has almost nothing to negotiate. The moment a deal grows enough moving parts to require real negotiation, the instrument that was supposed to save time starts costing it. For a structured walk through the instrument-selection decision, see Raising Capital Legally: A Comprehensive Guide to Securities Exemptions for Startups; for the corporate housekeeping that must already be in place before any of this works, see Incorporating a Startup the Right Way: A Founder's Legal Checklist.

A worked comparison. Suppose Meridian Robotics, Inc., a two-founder Delaware C-corp, needs $600,000 to reach a hardware prototype. Three angels are ready. Angel A is a former operator who has done a dozen SAFE deals and is indifferent. Angel B is a retired dentist investing for the first time who keeps asking "so it's a loan?" Angel C runs a small fund and wants downside protection. If Meridian issues SAFEs, Angel B's confusion is a real cost, and Angel C may push for a convertible note with a maturity date so that, if no priced round materializes, the debt gives leverage to renegotiate. Meridian could accommodate everyone with a single SAFE round and a short investor letter explaining the mechanics to Angel B, or it could bow to Angel C and run a note round with a 24-month maturity and 5% simple interest. The point is not that one answer is correct; it is that Meridian should choose consciously, having priced the friction, rather than defaulting to a SAFE because the template was already open in another tab.

Phase 2 — Set the four economic levers (and touch nothing else)

• [ ] Negotiate the valuation cap — the single most important early-stage term — which sets the maximum company valuation used to price the investor's conversion.
• [ ] Negotiate the discount (commonly 10–20%, with 20% the most frequent), which converts the SAFE at a set percentage below the priced-round price.
• [ ] Decide whether the SAFE carries a most-favored-nation (MFN) provision, letting the holder claim better terms later granted to another SAFE investor.
• [ ] Set the trigger — the dollar threshold of the future "Equity Financing" that causes conversion — and draft it so the threshold reflects a genuine institutional round, not a friends-and-family top-up.
• [ ] Use one current, standard form for every investor; push anything bespoke into a tightly scoped side letter, never into the body of the SAFE.
• [ ] Confirm which cap-and-discount combination applies: most SAFEs use a cap or a discount, some use both (with the investor taking whichever yields more shares), and a few use neither (an uncapped MFN SAFE).

Why discipline is the whole game. Beyond the dollar amount invested, a standard SAFE has essentially four levers, and a healthy one pulls only one or two. The instrument's economy is its point. Bolt on liquidation mechanics, redemption rights, or governance terms and you have not improved the SAFE; you have built a bad convertible note that also lacks a note's protections. Keep the negotiation to the cap and the discount, and the round stays a two-afternoon exercise.

The valuation cap does the heavy lifting. It answers the early investor's central fear: I am taking the most risk, at the earliest moment, and I do not want a later priced round's high valuation to price me in as if I had arrived late. The cap fixes a ceiling. If Meridian's SAFE carries a $6 million post-money cap and the Series A prices the company at $18 million pre-money, the SAFE investor converts as though the company were worth $6 million, capturing three times the shares per dollar that the new money receives. The cap is what makes writing a check into a pre-revenue robotics company rational; without it, the earliest and bravest capital would be punished for the company's own success. Seed caps vary enormously with sector, geography, and traction, but a rough band runs from the low single-digit millions to the high single digits or low teens. The number is a negotiation, not a science.

The discount is the cap's humble cousin. Where a cap is absolute, a discount is relative: it converts the SAFE at, say, 80% of the price the Series A investors pay, regardless of the valuation. A 20% discount is the modal term. When a SAFE has both a cap and a discount, the standard forms give the investor the benefit of whichever produces more shares — the cap governs in a strong up-round, the discount governs in a modest one. Founders sometimes fixate on the discount because the percentage is legible, but at typical seed valuations the cap almost always dominates the outcome. The discount is the safety net for the case where the priced round comes in below the cap.

The MFN clause is a ratchet on terms rather than price. It lets an early SAFE holder, when the company later sells a SAFE on more favorable terms, elect to adopt those terms retroactively. Y Combinator publishes a standalone "MFN SAFE" with no cap and no discount, relying entirely on this clause; it is the instrument of the investor who trusts the founder enough to defer the pricing question but wants assurance of never being the worst-treated party. MFN is investor-protective and should be granted with eyes open, because it means the terms you give the next investor silently reprice the terms you gave the last one.

The trigger is the least glamorous and most misdrafted lever. The SAFE converts on a qualifying "Equity Financing," typically defined as the company's sale of preferred stock above some threshold — often around $1 million — for the principal purpose of raising capital. Draft the threshold too low and a small friends-and-family preferred top-up can trip conversion at a valuation nobody intended to price against; draft the definition loosely and disputes arise over whether a given financing "counts." The threshold should describe a real institutional round. Watch, too, the SAFE's other exit paths: conversion on a liquidity event (acquisition or IPO) and on dissolution, each with its own payout formula that the parties should actually read rather than assume.

A cap in action. Return to Meridian. Suppose Angel A invests $200,000 on a SAFE with a $6 million post-money cap and no discount. Eighteen months on, a fund leads a Series A at a $24 million post-money valuation. Because the cap is $6 million, Angel A converts at a per-share price one-quarter of what the new investor pays. The $200,000 buys roughly the ownership that $800,000 of new money would buy at the round price. That four-to-one leverage is the cap doing its job — and, as Phase 4 will show, it is also the seed of the phantom-preference problem, because Angel A now holds cheap preferred stock that may carry a liquidation claim far larger than the $200,000 actually invested.

Phase 3 — Understand post-money versus pre-money and model the dilution

• [ ] Identify whether your form is pre-money (the early investor's percentage floats and absorbs later dilution) or post-money (the 2018 Y Combinator default, which fixes and protects the early investor's percentage).
• [ ] Recognize that under post-money SAFEs the protected percentages stack, and the resulting dilution lands on the founders and option pool, not on the SAFE holders.
• [ ] For each SAFE, estimate the locked-in ownership as roughly the investment amount divided by the post-money cap, then add every SAFE together.
• [ ] Model the cumulative effect of every tranche — the "SAFE pileup" — before selling the next one.
• [ ] Model the Series A conversion from the outset, including the new-hire option pool the priced round will demand, since the pool usually comes out of the pre-money and dilutes founders further.

Why the post-money shift was seismic. In 2018, Y Combinator replaced its original pre-money SAFE with a post-money version, and the change looked like plumbing. It was not. Under the old pre-money SAFE, the investor's ultimate percentage was uncertain, because it was computed on a pre-money capitalization that did not yet include the other SAFEs; every additional SAFE the company sold diluted the earlier SAFE holders along with the founders. Under the post-money SAFE, the investor's percentage is fixed at signing — computed as a share of the fully diluted capitalization including all the SAFE money — and it is protected against dilution from later SAFEs and notes. The clarity is genuinely better for everyone's modeling. But clarity has a distributional consequence: if the SAFE holders' percentages are locked and protected, then all the dilution from additional SAFEs must fall on the only parties left standing, which is the founders (and the employees whose options share the common pool).

The mechanism is easy to state and easy to underestimate. Each post-money SAFE carves out a fixed, protected slice of the company. Sell one $500,000 SAFE at a $5 million post-money cap and you have conveyed a protected 10%. Sell a second on the same terms and you have conveyed another protected 10% — and, crucially, that second slice does not dilute the first; it dilutes you. The percentages add. A founder thinking in dollars ("we've only raised a million") is not tracking the number that matters ("we've conveyed a protected twenty percent"). This is the SAFE pileup, and it is the mechanism behind nearly every "how did I end up with less than half my company?" story at Series A.

Worked math for the pileup. Lumen Health, Inc. starts with two founders splitting 8,000,000 shares of common, plus a 2,000,000-share option pool, for 10,000,000 fully diluted shares. Over a year, the founders raise on post-money SAFEs, all with a $10 million post-money cap:

• SAFE 1: $1,000,000 → protected 10%
• SAFE 2: $750,000 → protected 7.5%
• SAFE 3: $750,000 → protected 7.5%
• SAFE 4: $500,000 → protected 5%

That is $3,000,000 raised and 30% of the company conveyed to SAFE holders on a fully diluted, post-money basis, before the Series A. Now the priced round arrives. The lead wants to buy 20% for new money and insists the company expand the option pool to 15% of the post-financing cap table, with the pool created out of the pre-money (the standard ask). Do the arithmetic and the SAFE holders keep their protected 30%, the new investor takes 20%, the refreshed pool takes its share, and the founders — who began owning 80% of a 10-million-share company — walk out somewhere in the mid-40s to low-50s percent range, entirely from instruments they signed as routine seed paper. Nothing went wrong. Every document performed. The founders simply never ran the combined model.

The defense is unglamorous and completely effective: build one pro forma that lays every SAFE, the expected Series A size and valuation, and the new option pool on a single page, and get every party to sign off on that page before selling the next SAFE. The moment the founders can see the pileup as a percentage, they can price it, cap the number of SAFEs, or switch to a priced round while there is still room to choose. For the mechanics of how the conversion actually runs at the priced round — the denominator conventions, the option-pool shuffle, the down-round fiduciary scrutiny — work through Bringing On New Investors After the Seed Round: Navigating the First Priced Financing.

Phase 4 — Watch the phantom liquidation preference and run the conversion math

• [ ] Recognize that a SAFE converting at a cap or a discount receives preferred stock at a lower price per share but usually inherits the full Series A liquidation preference (typically 1x the round's original issue price).
• [ ] Compute the implied preference multiple: a 20% discount alone yields about 1.25x; a low cap against a high priced-round valuation can produce 3x, 4x, or more.
• [ ] Where the gap is large, negotiate the fix in the SAFE conversion mechanics — "shadow preferred" (a separate series whose preference matches the SAFE's lower purchase price) or "discount-in-common" (delivering the discount as extra common shares rather than cheap preferred).
• [ ] Fix the fully diluted denominator conventions in advance: whether converting SAFE shares and the new option pool sit in the pre-money or the post-money capitalization changes everyone's percentages.
• [ ] Build one pro forma model that every party reviews before closing. Never run conversions on a napkin at the signing table.

What a liquidation preference is, in one breath. When a company is sold, preferred stockholders are paid a fixed amount off the top before the common stockholders — the founders and employees — see anything. A "1x non-participating" preference means each preferred share gets back its original purchase price (or converts to common and shares pro rata, whichever is greater). The preference is measured by the original issue price of the preferred stock, and there is the trap.

The phantom multiple. When a SAFE converts, it typically converts into the same series of preferred stock the Series A investors buy — but at the SAFE's lower conversion price. The Series A investor pays, say, $2.00 per share and gets a 1x preference of $2.00 per share. The SAFE holder, converting at a $6 million cap against a $24 million valuation, might receive that same series at an effective $0.50 per share. If the SAFE holder inherits the stated 1x preference of the round — $2.00 per share — then the SAFE holder has paid $0.50 and holds a liquidation claim of $2.00. That is a 4x preference, invisible on the face of the documents, sitting quietly in the certificate of designation. A 20% discount produces a gentler version of the same effect: pay $1.60, hold a $2.00 preference, a 1.25x. The steeper the gap between a low cap and a high round price, the fatter the phantom multiple.

Why it matters is a matter of arithmetic and timing. In a blockbuster exit, nobody notices the phantom preference, because everyone converts to common and shares the upside; the preference is irrelevant when the pie is enormous. The phantom preference bites in the modest exit — the acquihire, the soft landing, the $30 million sale of a company that raised $8 million — precisely the outcome where every dollar off the top is contested and where the founders most need their common stock to be worth something. A 4x phantom preference on the SAFE money can hoover up the modest exit's proceeds before the common sees a cent.

The two standard fixes. Sophisticated forms solve this in one of two ways. Shadow preferred creates a separate sub-series (often "Series A-1") for the converting SAFEs, identical to the Series A in every respect except that its per-share liquidation preference equals the SAFE's actual, lower conversion price. The SAFE holder gets the cheap shares and the matching cheap preference — economically honest. Discount-in-common takes a different route: it delivers the cap-or-discount benefit as additional shares of common stock rather than as underpriced preferred, so the SAFE holder's preference tracks only the money actually invested. Both approaches neutralize the phantom multiple; both require the drafters to have thought about it before conversion. The widely used Y Combinator and Practical Law SAFE forms do not solve this for you as a default — the standard post-money SAFE converts into the priced-round preferred and inherits its preference — so a Series A lawyer who spots a fat phantom preference on a cap-heavy SAFE will not be shy about repricing the whole deal to claw it back. Better to have chosen the shadow-preferred structure at the seed than to negotiate it under duress at the A.

Denominator discipline. Two people can model the same round and reach different founder percentages, honestly, because they made different assumptions about what sits "in the pre-money." Are the converting SAFE shares counted before or after the new money? Is the enlarged option pool carved out of the pre-money (diluting founders and SAFEs) or the post-money (diluting everyone)? These are conventions, not laws of nature, and the term sheet should state them explicitly. The pro forma is the referee. One model, one page, agreed denominators, circulated before the closing — that single artifact prevents more Series A heartburn than any clever drafting.

Phase 5 — Comply with the securities laws

• [ ] Treat every SAFE sale as the sale of a security, and identify the specific exemption you are relying on before you accept a dollar.
• [ ] Default to Regulation D, Rule 506(b): unlimited accredited investors plus up to 35 sophisticated non-accredited purchasers, with no general solicitation or advertising.
• [ ] Confirm each investor's accredited status under Rule 501, using the income, net-worth, or newer professional-credential and knowledgeable-employee categories.
• [ ] If you must solicit publicly, use Rule 506(c) instead and take reasonable steps to verify accreditation — self-certification alone will not do.
• [ ] Screen for "bad actor" disqualification under Rule 506(d) across the issuer, its directors, officers, 20% owners, and any paid solicitors.
• [ ] File a Form D with the SEC within 15 days of the first sale, and make the state "blue sky" notice filings and pay the fees each covered state requires.
• [ ] Respect the anti-fraud rules — Securities Act § 17(a), Exchange Act § 10(b), and SEC Rule 10b-5 — which apply no matter which exemption you use.
• [ ] Watch integration under Rule 152 if a SAFE round sits close in time to another offering.

Why none of this is optional. Section 5 of the Securities Act of 1933 makes it unlawful to offer or sell any security in interstate commerce unless a registration statement is in effect or an exemption applies. 15 U.S.C. § 77e. A SAFE is unquestionably a security — it is an investment contract, the paradigm case since SEC v. W.J. Howey Co., 328 U.S. 293 (1946), of money invested in a common enterprise with profits expected from others' efforts. Startups never register; registration is a public-offering apparatus that costs millions and takes months. So the entire round lives or dies on the exemption. Get it wrong and the consequences are not academic: the investors acquire a right of rescission (they can demand their money back, often after the money is spent), and the company and its principals face potential SEC enforcement and state action.

Section 4(a)(2) and the private-offering baseline. The foundational exemption is Section 4(a)(2), which exempts "transactions by an issuer not involving any public offering." 15 U.S.C. § 77d(a)(2). The Supreme Court gave that phrase its enduring gloss in SEC v. Ralston Purina Co., 346 U.S. 119 (1953), holding that the exemption turns on whether the offerees "are shown to be able to fend for themselves" — whether they have access to the kind of information registration would provide and the sophistication to evaluate it. Ralston Purina is elegant but fuzzy: it offers a standard, not a rule, and issuers dislike betting a whole round on an after-the-fact judgment about offeree sophistication. The leading treatises — Louis Loss, Joel Seligman & Troy Paredes, Securities Regulation, and Thomas Lee Hazen, The Law of Securities Regulation — trace decades of litigation over where the line falls. That uncertainty is exactly why Regulation D exists.

Regulation D as the safe harbor. Regulation D, and specifically Rule 506, is a non-exclusive safe harbor that converts the Ralston Purina standard into workable bright lines. Two doors matter for SAFE rounds:

Rule 506(b) — the workhorse. It permits an issuer to raise an unlimited amount from an unlimited number of accredited investors, plus up to 35 non-accredited purchasers who are financially sophisticated (or advised by someone who is), provided the issuer engages in no general solicitation or general advertising. 17 C.F.R. § 230.506(b); see also the general-solicitation prohibition at 17 C.F.R. § 230.502(c). In practice, most startups sell only to accredited investors under 506(b) to avoid the enhanced disclosure obligations that kick in the moment a single non-accredited investor participates. The catch of 506(b) is the solicitation ban, discussed below.

Rule 506(c) — the public-solicitation door. Added by the JOBS Act reforms effective September 23, 2013, Rule 506(c) permits general solicitation and advertising — you may tweet the round, post it on a platform, pitch it from a stage — but only if every purchaser is in fact accredited and the issuer takes "reasonable steps to verify" that accredited status. 17 C.F.R. § 230.506(c). Verification is the price of publicity, and it is a real burden. The SEC was explicit that a box-checked self-certification does not cut it: an issuer "will [not] have taken reasonable steps to verify accredited investor status if it, or those acting on its behalf, required only that a person check a box in a questionnaire or sign a form." The rule supplies non-exclusive verification methods — reviewing two years of IRS income forms, reviewing bank and brokerage statements for net worth, or obtaining a written confirmation from the investor's lawyer, CPA, broker-dealer, or registered investment adviser — and a 2020 amendment added a fifth method letting an issuer rely on a prior verification within the past five years plus the investor's re-certification. 17 C.F.R. § 230.506(c)(2)(ii).

Who is accredited. Rule 501 defines the accredited investor. 17 C.F.R. § 230.501. For a natural person, the classic tests are income — over $200,000 individually (or $300,000 jointly with a spouse or spousal equivalent) in each of the two most recent years, with a reasonable expectation of the same in the current year — or net worth exceeding $1,000,000, excluding the value of the person's primary residence. The SEC's 2020 amendments meaningfully widened the door: individuals now also qualify by holding certain professional credentials (the Series 7, Series 65, and Series 82 licenses), and "knowledgeable employees" of a private fund qualify as to that fund regardless of income or net worth. Entities qualify through their own asset tests or, under Rule 501(a)(8), where all of the entity's equity owners are themselves accredited — a look-through that matters when angels invest through an LLC. Getting this right is not paperwork for its own sake; the exemption depends on it, and a signed accredited-investor representation is the standard evidence that the issuer did its diligence.

The general-solicitation trap. The 506(b) solicitation ban catches founders constantly, because the modern fundraising world is built on broadcasting. A demo-day pitch to a room of strangers, a LinkedIn post announcing "we're raising," a public listing on an online fundraising platform, an email blast to a purchased list — each can constitute general solicitation and blow a 506(b) exemption, converting the round into an unregistered public offering. The SEC's touchstone is a "pre-existing, substantive relationship" between the issuer (or a broker acting for it) and the offeree; cold public outreach fails that test. The practical rule for a 506(b) SAFE round: raise from people you or your investors actually know, keep the round off the open internet, and if you genuinely need to solicit the public, switch deliberately to 506(c) and accept the verification burden. Do not straddle the two.

Bad actors. Rule 506(d) disqualifies an offering from the safe harbor entirely if any "covered person" — the issuer, its directors, executive officers, 20%-or-greater beneficial owners, and any compensated solicitors — has a disqualifying event in their past, such as certain securities-related criminal convictions, court injunctions, or regulatory bars. 17 C.F.R. § 230.506(d). There is a reasonable-care exception where the issuer, despite a factual inquiry tailored to the offering, could not have known of the disqualification, and pre-September 23, 2013 events do not disqualify but must be disclosed to purchasers under Rule 506(e). The operational fix is a bad-actor questionnaire circulated to every covered person before the first sale — cheap insurance against an expensive surprise.

Form D and blue sky. Filing a Form D with the SEC is not what perfects the exemption — the exemption exists independently — but Rule 503 requires the notice within 15 days of the first sale of securities in the offering, and missing it can jeopardize the issuer's ability to rely on Regulation D in some states and invites regulatory attention. On the state side, the National Securities Markets Improvement Act of 1996 (NSMIA) amended Section 18 of the Securities Act to make Rule 506 securities "covered securities," preempting state registration requirements. 15 U.S.C. § 77r. But preemption is partial: states may still require notice filings and collect fees for offerings of covered securities sold into their borders, and they retain their anti-fraud authority. So a SAFE round sold to investors in five states still generally means one federal Form D plus five state notice filings and five sets of fees. For the mechanics of that state-by-state patchwork under the Uniform Securities Act regimes, the blue-sky analysis is genuinely fiddly and jurisdiction-specific; budget for it.

The anti-fraud rules never sleep. Whatever exemption you claim, the anti-fraud provisions apply to every securities transaction: Section 17(a) of the Securities Act (15 U.S.C. § 77q(a)), Section 10(b) of the Exchange Act (15 U.S.C. § 78j(b)), and SEC Rule 10b-5 (17 C.F.R. § 240.10b-5) prohibit material misstatements and omissions in the offer or sale of a security. Exemption from registration is not exemption from honesty. A founder who oversells the pipeline, hides a co-founder's departure, or papers over a threatened lawsuit in the SAFE pitch can face rescission and fraud liability no matter how clean the Rule 506 paperwork looks. The private-placement discipline of accurate, complete disclosure is not a courtesy to investors; it is a defense to the company.

Integration and adjacent offerings. If the SAFE round sits close in time to another capital raise — a simultaneous crowdfunding campaign, a later priced round, a separate 506(c) tranche — the doctrine of integration asks whether the offerings should be treated as one, which could contaminate an exemption (for example, by combining a publicly solicited 506(c) tranche with a quiet 506(b) tranche). The SEC's 2021 reforms replaced the old five-factor integration test with a more issuer-friendly framework and safe harbors under Rule 152, including a general 30-day safe harbor between distinct offerings. 17 C.F.R. § 230.152. It is rarely a problem for a plain SAFE round, but it becomes one the moment a founder runs two differently-exempt raises in overlapping windows without advice.

Doing this without counsel is a false economy. The recurring theme of Phase 5 is that a SAFE is genuinely simple to draft and genuinely unsimple to sell lawfully. The overview of the whole exemption architecture — Rule 504, Regulation A+, and Regulation Crowdfunding under Section 4(a)(6), each an alternative for certain rounds — lives in Raising Capital Legally: A Comprehensive Guide to Securities Exemptions for Startups. A founder who cannot say which subsection of Rule 506 the round relies on is a founder who should not yet be accepting checks.

Phase 6 — Handle side letters, pro rata rights, and information rights

• [ ] Limit side letters ruthlessly; every promise in one is a separate obligation to track, honor, and disclose in future diligence.
• [ ] Grant pro rata (preemptive) rights to as few investors as possible, and where pressed, cap the right to the Next Equity Financing rather than every future round in perpetuity.
• [ ] Decline board seats, board-observer rights, and information rights for seed investors absent a genuine, specific reason.
• [ ] Remember that an MFN clause turns the next investor's side letter into a repricing of the last investor's terms; model the ripple before you sign.
• [ ] Keep a live schedule of every side letter, cross-referenced to the cap table, so nothing is forgotten when the Series A diligence request arrives.

Why side letters quietly recreate the problem the SAFE was meant to solve. A pro rata right lets an investor maintain its percentage by buying its share of a future round. It sounds harmless and often is, but it has gravity: once one investor gets a side letter, the next asks why she cannot have one too, and a stack of overlapping pro rata and information promises reconstitutes exactly the negotiation overhead the SAFE was supposed to eliminate. Worse, an unbounded pro rata right — "the right to participate in all future financings" — can complicate every subsequent round for the life of the company, because the company must offer participation and paper waivers each time. The disciplined move is to grant pro rata only to lead or strategically important investors, and to scope it to the Next Equity Financing, extinguishing after the Series A closes.

Board seats and information rights deserve a firmer no at the seed stage. A SAFE holder is not yet a stockholder; giving a seed investor a board seat or granting broad information rights before the priced round hands away governance and confidentiality for capital that has not even converted. Save those for the investors who lead priced rounds and earn them. The MFN interaction is the subtle trap: if early SAFEs carry MFN and you grant a later SAFE holder a pro rata right or a better cap, the MFN holders may claim the same, quietly propagating the concession across the whole seed stack. Every side letter is therefore not a bilateral deal but a potential amendment to the entire round. Track them as such. As new investors arrive across successive rounds and the obligations compound, keep the cumulative picture honest with the workflow in Bringing On New Investors After the Seed Round.

Phase 7 — Manage the cap table, the tax picture, and clean closing mechanics

• [ ] Record each SAFE on the cap table using its valuation cap as the modeling stand-in, with footnoted assumptions, and note explicitly that a lower priced-round valuation converts the SAFE into more shares.
• [ ] Avoid a vague "Friends & Family" or "SAFEs" line item; model each instrument with its own stated assumptions so the pileup is visible.
• [ ] Confirm the SAFE's effect on the common stock's 409A valuation and option strike prices, and refresh the 409A when the SAFE round materially changes the capitalization.
• [ ] Flag the unsettled federal tax treatment of SAFEs with tax counsel, including the qualified small business stock (QSBS) holding-period question under Section 1202.
• [ ] Collect signed accredited-investor representations, the bad-actor questionnaires, and any side letters into the closing binder, and calendar the Form D and blue-sky filing deadlines.
• [ ] Run the intended SAFE terms and every side letter past securities and tax counsel before signature — not after.

The cap table is a modeling document, not a ledger. Because a SAFE converts into an unknown number of shares (the number depends on a future priced-round valuation that does not yet exist), the honest way to show it is a modeled line using the cap as the conversion assumption, with a footnote stating the assumption and the counterintuitive direction of the risk: the worse the company does — the lower its Series A valuation — the more shares the SAFE converts into, because a lower price per share means more shares to deliver the fixed dollar or fixed percentage. Founders who lump SAFEs into a fuzzy block on the cap table are the founders who get surprised. The clean, footnoted, assumption-stated model is the single artifact that keeps a founder from being ambushed by her own paperwork. Getting this house in order before the raise is a core part of being diligence-ready; see Getting Diligence-Ready Before You Raise: A Comprehensive Guide for Founders.

409A and cheap options. One quiet advantage of SAFEs and notes over a priced round is that they tend to keep the fair market value of the common stock — and therefore the strike price of employee options — lower. Under Section 409A of the Internal Revenue Code, a stock option is exempt from that section's punishing deferred-compensation regime only if its exercise price is at least the fair market value of the underlying common stock on the grant date. 26 U.S.C. § 409A; 26 C.F.R. § 1.409A-1(b)(5). Set the strike below fair market value and the option holder can face a 20% additional tax plus interest on the spread — a catastrophe for the very employees the options were meant to reward. A priced preferred round sets an explicit valuation that pulls the common's fair market value up; a SAFE round, structured as deferred equity, exerts less upward pressure, which is why companies typically obtain an independent 409A appraisal (the safe-harbor method that carries a presumption of reasonableness) and refresh it when a financing materially changes the capitalization. The upshot: after a SAFE round, get or update the 409A before granting the next batch of options.

The tax cloud over the SAFE itself. The federal income tax treatment of a SAFE is genuinely unsettled, and honest tax counsel will say so. The instrument is neither plainly debt nor plainly equity: various practitioners characterize it as a prepaid forward contract, a variable prepaid forward, or an equity interest from issuance, and the IRS has issued no definitive guidance. The characterization is not academic. If a SAFE is not treated as stock until conversion, then the five-year holding period for qualified small business stock under Section 1202 — which can exclude a large share of gain from federal tax on a later sale — generally does not begin until the SAFE converts into actual preferred stock at the Series A, potentially costing an investor years of holding-period credit compared with buying stock directly. 26 U.S.C. § 1202. That is a material investor consideration and a reason sophisticated angels sometimes prefer a priced round or a note structured to start the clock earlier. Flag it; do not resolve it on a founder's intuition.

Closing mechanics. A SAFE closing is mercifully light, but "light" is not "none." Assemble the executed SAFEs, the signed accredited-investor representations, and the completed bad-actor questionnaires into a closing binder; confirm the board has authorized the issuance (a short unanimous written consent) and, if the SAFE promises to reserve shares, that the authorized-but-unissued preferred exists or can be created; calendar the Form D deadline (15 days from the first sale) and the blue-sky notice filings; and update the cap table and the company's data room the same week, while the terms are fresh. The instruments that get lost are the ones nobody filed away, and a missing SAFE surfacing during Series A diligence is a self-inflicted wound that erodes investor confidence at the worst possible moment.

Common failure modes

• Reaching for a SAFE reflexively. The instrument should fit a considered strategy, not stand in for the absence of one. If the deal wants three side letters, it wants a priced round.
• Negotiating bespoke economics into the SAFE body. Keep to the four levers; a SAFE with a custom liquidation waterfall is a defective convertible note.
• Ignoring post-money stacking. Protected percentages add up and land on the founders. The number that matters is percentage conveyed, not dollars raised.
• Skipping the conversion model. The dilution surprise almost always visits the team that treated conversion as tomorrow's problem.
• Overlooking the phantom liquidation preference. A low cap or a steep discount can hand the SAFE holders an off-market multiple that only appears in a modest exit — the exit where it hurts most.
• Treating the SAFE as outside the securities laws. It is a security. Pick your Rule 506 door, verify accreditation, screen for bad actors, file the Form D, and never generally solicit under 506(b).
• Letting the general-solicitation line blur. A single public "we're raising" post can convert a quiet 506(b) round into an unregistered public offering.
• A vague "Friends & Family" block on the cap table. Model each SAFE with stated assumptions, or the pileup stays invisible until it is irreversible.
• Forgetting the state filings. NSMIA preempts state registration for Rule 506 offerings, not state notice filings and fees. Five states of investors means five filings.

Primary and secondary authority

Registration and the private-offering exemption

• Securities Act of 1933 § 5, 15 U.S.C. § 77e — registration requirement for any offer or sale of a security.
• Securities Act of 1933 § 4(a)(2), 15 U.S.C. § 77d(a)(2) — private-offering exemption, construed in SEC v. Ralston Purina Co., 346 U.S. 119 (1953) (offerees able to "fend for themselves").
• SEC v. W.J. Howey Co., 328 U.S. 293 (1946) — the investment-contract test that makes a SAFE a security.

Regulation D

• Rule 501, 17 C.F.R. § 230.501 — accredited-investor definitions (income, net worth, professional credentials, knowledgeable employees, all-accredited entities).
• Rule 502(c), 17 C.F.R. § 230.502(c) — general-solicitation prohibition.
• Rule 503, 17 C.F.R. § 230.503 — Form D notice, due within 15 days of first sale.
• Rule 506(b), 17 C.F.R. § 230.506(b) — unlimited accredited plus up to 35 sophisticated non-accredited purchasers, no general solicitation.
• Rule 506(c), 17 C.F.R. § 230.506(c) — general solicitation permitted with reasonable-steps verification of accredited status.
• Rule 506(d)–(e), 17 C.F.R. § 230.506(d)–(e) — bad-actor disqualification and disclosure of pre-2013 events.
• Rule 152, 17 C.F.R. § 230.152 — integration framework and safe harbors (2021 reforms).

State law and anti-fraud

• National Securities Markets Improvement Act of 1996; Securities Act § 18, 15 U.S.C. § 77r — "covered securities" preemption of state registration, preserving state notice filings and fees.
• Uniform Securities Act (1956 and 2002) — the models behind most state blue-sky statutes.
• Securities Act § 17(a), 15 U.S.C. § 77q(a); Securities Exchange Act § 10(b), 15 U.S.C. § 78j(b); SEC Rule 10b-5, 17 C.F.R. § 240.10b-5 — anti-fraud rules that apply regardless of exemption.

Tax

• Internal Revenue Code § 409A, 26 U.S.C. § 409A; 26 C.F.R. § 1.409A-1(b)(5) — option strike price must equal or exceed common-stock fair market value.
• Internal Revenue Code § 1202, 26 U.S.C. § 1202 — qualified small business stock exclusion and its five-year holding period, relevant to a SAFE's uncertain start date.

Secondary sources

• Louis Loss, Joel Seligman & Troy Paredes, Securities Regulation (multivolume treatise).
• Thomas Lee Hazen, The Law of Securities Regulation.
• Practical Law, Startup Seed Financing Instruments: Convertible Notes and SAFEs, and the Choosing the Right Seed Financing Instrument for Startups Checklist.
• Y Combinator, Post-Money SAFE (2018) and Safe Financing Documents User Guide.

Related UpLaw Resources

Related Guides and Overviews

• Simple Agreements for Future Equity: An Overview of How SAFEs Actually Work
• Raising Capital Legally: A Comprehensive Guide to Securities Exemptions for Startups
• Bringing On New Investors After the Seed Round: Navigating the First Priced Financing
• Getting Diligence-Ready Before You Raise: A Comprehensive Guide for Founders
• The Startup Legal Document Stack: What Founders Need and Why

Related Checklists

• Incorporating a Startup the Right Way: A Founder's Legal Checklist

Related Toolkits

• The Founder's Path from Formation Through Financing: A Roadmap and Resource Toolkit

Related Articles

• Employee Invention Assignment Agreements: Securing Clean Title Across Jurisdictions
• Running Multiple Businesses Under One Structure: A Deep Dive into Holding, Operating, and Parent Companies
• How Cryptocurrency Is Regulated Around the World: A Comparative Overview
• Types of Lawyers: A Field Guide to Legal Specialties and Choosing the Right One

This checklist provides general information for research and is not legal advice. A SAFE is a security; its use, conversion mechanics, tax treatment, and securities-law compliance vary by jurisdiction and by the facts of a given financing. Consult qualified securities counsel and a startup-experienced accountant before issuing or investing in a SAFE.