NFL guaranteed money explained: what is actually locked in

Every large contract in the sport is announced with two numbers. There is the total, which is close to fiction, and there is the guaranteed figure, which everybody treats as the real one. The second number is better than the first. It is still not what the player will be paid, and the distance between the two is where NFL guaranteed money actually lives.

A guarantee in this sport is not a single thing. It is a bundle of separate promises, written against separate events, attaching on separate dates, backed by a bank deposit that most supporters have never heard of and reduced by a clause that agents will hold a signing hostage over. Two contracts with identical announced guarantees can be worth wildly different amounts of cash, and the difference is visible in the paperwork rather than in the press release.

None of this is hidden. The mechanisms sit in the collective bargaining agreement and in the standard player contract appended to it, both of which are published. They just get compressed, in the retelling, into one word that hides five decisions.

This is those five decisions, in the order they matter.

What NFL guaranteed money actually is, and why it is written as a waiver

Start with the thing that makes the whole system make sense: a guarantee does not add an obligation to the contract. It removes an escape from it.

The standard player contract, the form every deal in the league is built on top of, already says the club owes the player a salary. That salary sits in paragraph 5, which is why people in the industry say "Paragraph 5 salary" when they mean base pay. If nothing else were written, the money would simply be owed.

What makes the money conditional are the paragraphs that follow. One of them lets a club terminate the contract if, in its own judgement, the player's skill or performance has been unsatisfactory compared with others competing for a place on the roster, or if his conduct reflects badly on the club, or if he is expected to contribute less than somebody the club wants to sign in his place while a cap is in force. Another deals with injury and sets out what happens when a player cannot pass a physical. Those clauses, not the absence of a promise, are what make the contract cancellable.

A guarantee is therefore a negotiated switch-off. The parties agree that for a specified sum, in specified years, the club gives up the right to use a specified escape. Everything else about the way guarantees are described follows from that structure, including the odd fact that a guarantee has to name which escape it blocks.

• 3Termination grounds a guarantee can be written against
• 15Funding rule deductible per club, $m, 2020 to 2028
• 17Funding rule deductible per club, $m, from 2029
• 75Ceiling on the share of a contract counted as fundable pay, %

All four are set by the collective bargaining agreement rather than by any individual deal. The deductible under the funding rule applies once per club rather than once per contract.

That is why "guaranteed" on its own is an incomplete sentence in this sport. Guaranteed against what?

The three protections, and why only one of them is worth much alone

Money in an NFL contract can be protected for skill, for injury, and for cap. Those are the three, and a contract can carry any combination of them on any portion of the salary in any year. The combinations are where the real negotiation happens.

Injury protection is the narrow one and the one players want most, because injury is the risk they cannot manage by playing better. If the player is hurt performing his job, cannot pass the club physical, and is released while in that condition, the protected salary is still owed. That is genuine and it is not nothing.

It also has a trapdoor that catches people out every summer. The protection is triggered by the state the player is in when the contract ends, not by the fact that he was once injured. A player who tears something in October, rehabilitates through the spring, passes his physical in July and is released in August has not triggered an injury guarantee at all. He got healthy, which is exactly what everybody wanted, and getting healthy dissolved his protection.

Skill protection covers being beaten out. If the club releases the player because somebody younger is better, or the coordinator changed, or the scheme no longer has a role for him, the money is owed anyway. This is the protection clubs resist hardest, and for an obvious reason: it is the one that stops them correcting an evaluation error. Every general manager is wrong about players. Skill protection makes being wrong expensive.

Cap protection covers termination for roster accounting reasons rather than football ones. On its own it is nearly decorative. A club that wants a player gone can almost always articulate a football reason, and the termination clause invites it to do so in its own judgement. Cap protection matters as the third leg of a stool, not as a stool.

Stack all three on the same dollars in the same year and the salary is fully guaranteed. That phrase is not a fourth, superior category. It is shorthand for the club having given up every route out at once, which is precisely what supporters assume a contract in any other sport already means.

The most common partial guarantee is injury alone, and it is common because it costs the club the least. Injury-only money is protection against an event with a known frequency that clubs are already insuring against in other ways. Skill money is protection against the club's own judgement, which no insurer will write a policy on.

Guaranteed at signing is the only phrase that means anything in March

The second variable is timing, and it does more damage to the headline figure than the protections do.

Money guaranteed at signing is certain from the day the contract is executed. Money that becomes guaranteed later is not. A very large share of what gets reported as guarantee is the second kind: salary that carries injury protection now and converts to full protection on a named date in a future league year, but only if the player is still under contract when that date arrives.

Those dates are negotiated, not set by rule, and they cluster in the opening days of a league year in March because that is when the roster decisions have to be made anyway. The effect is a scheduled decision point. A club looking at a player whose next season vests on the fifth day of the league year is not asking whether he is worth the money over the life of the deal. It is asking whether he is worth it before Thursday.

Agents call an early vesting date player-friendly and a late one club-friendly, and the language is exactly right. The date is not administrative housekeeping. It is the moment the club has to either commit or move, and pushing it later in the calendar is worth real money to the club because it buys more information before the commitment binds.

Contracts written this way produce a rolling structure. Year two vests early in year two, year three vests early in year three, and so on down the deal. On paper the player has three guaranteed years. In practice he has one guaranteed year and two annual reviews, each with a hard deadline, each of which the club will approach with a spreadsheet showing what releasing him costs against what keeping him costs.

1. 1The protections are chosenThe parties decide which of skill, injury and cap each block of salary is protected against, and in which years. Money carrying all three is fully guaranteed. Money carrying one is not, whatever the announcement says.
2. 2The vesting dates are setSalary not guaranteed at signing is given a date on which its injury protection converts to full protection. The date is negotiated. Earlier favours the player, later favours the club.
3. 3The signing bonus is paidUsually within weeks. This portion is now beyond argument, because a release cannot recover money already in the player's account except in narrow forfeiture cases.
4. 4The club funds the restGuaranteed compensation not yet paid may have to be deposited into a segregated account, discounted to present value, after a club-wide deductible. This is a cash requirement, not a cap one.
5. 5The first vesting date arrivesThe club decides before the deadline. Keep him and the next tranche locks in. Release him first and that tranche was never guaranteed, though injury protection may still bite if he cannot pass a physical.
6. 6The salary is earned week by weekBase salary is paid across the regular season rather than in one payment, which is why a release in October costs a club less cash than the same release in March.
7. 7Offset language settles the balanceIf the player was released with guaranteed money outstanding and signs elsewhere, an offset clause lets the first club deduct the new salary from what it still owes. Without one, he collects from both.

The sequence is the same for every guaranteed contract in the league. What varies between deals is only how far down this list the money survives.

There is a third category that never appears in any announcement and matters as much as either of the first two. Call it the practical guarantee: salary that carries no protection on paper but which the club will pay anyway, because releasing the player would accelerate more dead charge onto the cap than keeping him costs. A contract structured with a large bonus early creates a stretch of years in the middle that are functionally locked in by arithmetic rather than by language. Working out where that stretch sits requires reading the proration schedule alongside the guarantee schedule, which is the sort of thing the accounting side of the cap governs rather than the contract itself.

Why NFL guaranteed money has to sit in a bank account before it is owed

Here is the mechanism that explains more about this sport's contracts than the salary cap does, and which almost never appears in the coverage.

Guaranteed compensation that a club has promised but not yet paid can be required to be funded. Not accrued, not provided for, not budgeted. Deposited, into a segregated account the club cannot touch, before the season in which it is owed.

The agreement sets it out with some precision. The league may require each club, by a prescribed date, to deposit the present value of deferred and guaranteed compensation it owes, calculated at a discount rate defined in the agreement, less a deductible. That deductible is fifteen million dollars for the 2020 to 2028 league years and rises to seventeen million for 2029 and 2030. For guaranteed contracts specifically, the unpaid compensation counted in the calculation is capped at seventy-five per cent of the total compensation in that contract. Salary payable under an injury guarantee does not enter the calculation at all until the club has acknowledged that the player's injury qualifies him to receive it.

Read that again with the deductible in mind, because the deductible is per club, not per contract. It is a single allowance applied across the whole book of guarantees the club is carrying. A club with one guaranteed deal barely notices the rule. A club that has guaranteed money to eight players has burned the allowance on the first two and is writing cheques for the rest.

Guaranteed compensation promised, not yet paid90m

Club deductible, applied once across the roster15m

Balance to be deposited, before discounting75m

The guarantee figures are invented to show the arithmetic. The fifteen million deductible, and its application once per club rather than once per contract, are set by the collective bargaining agreement. The deposited sum is discounted to present value, so the real figure is lower again.

Show the numbers

Invented example: what the funding rule asks a club to deposit

Item	Value
Guaranteed compensation promised, not yet paid	90m
Club deductible, applied once across the roster	15m
Balance to be deposited, before discounting	75m

The consequence is a cash-flow constraint sitting underneath a system everybody analyses as a cap constraint. Two clubs with identical cap space are not equally able to offer a four-year full guarantee, because one owner can put the money in the account tomorrow and the other cannot without borrowing against something. That asymmetry is invisible in every cap table published anywhere, and it is a large part of why the same handful of clubs keep appearing at the front of the guarantee market.

It also explains a pattern that otherwise looks like collective stubbornness. A club offering a huge total with a modest guarantee is not necessarily being stingy about the money. It is often being careful about the timing of the money, which is a different objection and one the player cannot argue away with production.

The rule was written for a reason that has aged badly. In an era when a club going under was a live possibility, a player owed money in three years' time needed something better than a promise, and the escrow was that something. The league it protects players in now is not that league. The players' association has been trying to get the rule loosened for years, on the grounds that it now works against the people it was designed to protect, which is a genuinely awkward argument to find yourself on the wrong side of.

There is a further twist worth knowing. In January 2025 an arbitrator dismissed a grievance alleging that clubs had colluded to resist fully guaranteed veteran contracts, while finding that the league's own executives had urged owners to reduce guarantees at a meeting the previous year. An appeal panel upheld the dismissal in April 2026 on the same reasoning: encouragement was established, agreement between clubs was not. Whatever else that outcome settles, it is the most detailed public account there has been of why the fully guaranteed deal has stayed rare, and it did not come out of a negotiation.

The signing bonus is the only guarantee nobody can argue about

Every other protection is a promise. The signing bonus is a transfer.

It is paid in cash, usually within weeks of the contract being executed, and once it is in the player's account no subsequent release recovers it. The club can terminate the deal the following morning and the bonus stays paid. That is a stronger position than any guarantee language can create, because language can be litigated and a completed payment cannot.

The exceptions are narrow and they concern conduct rather than performance. Bonus money can be made subject to forfeiture where a player breaches specified obligations, and the circumstances in which a club can actually recover paid bonus are tightly bounded. Nobody structures a contract around them.

This is why the signing bonus is the number an agent works hardest on, and why two deals with the same announced total and the same announced guarantee can be meaningfully different if one has twice the bonus. It is also why a club that wants to look generous without being generous will offer total and guarantee, and hold the line on bonus. A guarantee that vests in two years is a promise about the future. A bonus is a Tuesday.

Two relatives of the bonus behave differently and are worth separating. A roster bonus is paid for being on the roster on a stated date, and it functions as a deadline rather than as compensation: the club has to decide before the date whether the player is worth it. An option bonus is paid when the club exercises an option built into the contract, and it prorates for cap purposes the same way a signing bonus does, which makes it a restructure written into the deal before anybody needed one. The mechanics of turning salary into bonus after the fact belong to the restructure, and they change the cap position without changing the guarantee at all.

Offset language, and the clause that decides whether a player is paid twice

A released player with guaranteed money outstanding is in an unusual position. He is owed the balance by his old club, and he is free to sign with a new one. Whether he collects both depends on a single clause.

With offset language, the old club deducts what the new club pays him from what it still owes. If eight million remains guaranteed and he signs for three, the old club pays five and the new club pays three, and the player receives the eight he was promised, no more.

Without offset language, he receives all eight from the old club and all three from the new one. The old club's obligation is fixed and the player's employment elsewhere is his own business.

Agents fight over this clause with an energy that surprises people who assume the big numbers are the argument. For an established player with a real chance of being released and a real chance of being wanted afterwards, the absence of offsets is worth a material amount of money in a scenario that is far from hypothetical. For a player who will retire if released, it is worth nothing. Which is why the fight over offsets tells you what both sides privately expect to happen.

There is a subtler version that produces most of the stalling in rookie negotiations. Rather than accept or reject offsets outright, the parties argue about which years they apply to, and about whether an offset applies to the whole of the new salary or only the part above a stated floor. A partial offset in year four of a deal is a small clause with a small expected value, and it can still hold up a signing for weeks, because both sides know the precedent it sets for the next player in the same draft slot.

What happens to guaranteed money when the club releases him anyway

Guarantees do not stop releases. They price them.

When a club releases a player with guaranteed salary outstanding, the money is still owed and the cap has to recognise it now rather than later. Guaranteed salary for future years accelerates onto the current year's cap alongside any unamortised signing bonus, which is the sharpest reason a fully guaranteed multi-year deal is a heavier commitment than the equivalent cash spread over the same period. The cap does not let a club walk away slowly.

Injury changes the route. A player who is hurt cannot simply be released, and the process that follows is a separate piece of machinery involving the injured reserve list, a physical, and in many cases a negotiated injury settlement: a payment agreed between club and player that ends the contract early and frees him to sign elsewhere once a set period has passed. Settlements exist because carrying an injured player on a roster spot for a whole season is expensive in a way that has nothing to do with money, and the rules governing that list, and who may come back from it, are set out in the piece on how injured reserve actually works.

There is one further protection that operates outside the guarantee system entirely and that almost nobody counts. Termination pay is a benefit under the collective agreement rather than a contract term. A player who has earned four or more credited seasons, who was on the active or inactive list on or after the date of his club's first regular season game, and who is then released, is entitled to the balance of his salary for that season. He is entitled to it once in his career, with a limited second entitlement if the first payment was small. It is not negotiated, it does not appear in any announcement, and for a veteran on a modest one-year deal it can be worth more than everything his contract calls guaranteed.

25%12%18%45%

• Signing bonus, already paid25m
• Base salary fully guaranteed at signing12m
• Salary guaranteed for injury, vesting later18m
• Salary carrying no guarantee at all45m

A constructed contract used to show where an announced guarantee sits. The same headline figure can be composed very differently, and the composition is the difference between a strong deal and a weak one.

Show the numbers

Invented four-year deal, announced at $100m with $55m guaranteed

Item	Value
Signing bonus, already paid	25m
Base salary fully guaranteed at signing	12m
Salary guaranteed for injury, vesting later	18m
Salary carrying no guarantee at all	45m

Read the four slices in that figure against each other and the announcement stops being informative. Twenty-five million is certain because it has been handed over. Twelve is certain because all three protections attach immediately. Eighteen is a decision the club will take on a date in a future March, protected in the meantime only against an injury the player has to still be suffering from. And forty-five is a number that exists to make the total look like a hundred.

The player in that invented deal has been announced at fifty-five guaranteed. He is genuinely guaranteed thirty-seven.

Split contracts, and the pay cut that arrives with the injury

At the other end of the market there is a structure that runs the guarantee logic backwards.

A split contract names two salary figures: one if the player is on the active roster, and a lower one if he is on injured reserve. It is not a penalty and it is not hidden, it is agreed in advance, and it is offered to players with limited leverage, typically undrafted rookies and veterans with complicated medical histories. The club is buying the downside. If this player breaks, we pay less while he is broken.

For a player at the bottom of a roster the split is often the price of the roster spot itself, and refusing it means the offer goes to somebody who will accept it. That is a real bargain being struck by consenting parties. It is also the reason a headline about a player's salary can be wrong by a wide margin in the specific weeks when he most needs the money, and it sits awkwardly beside the fact that the injury which triggers the lower rate was sustained doing the job.

The collective agreement provides a floor underneath this in the form of injury protection benefits, which pay a player who is unable to play the following season because of an injury suffered under contract, subject to conditions on service time and on the injury having been properly reported. It is a benefit rather than a guarantee, it is capped, and it is claimed rather than paid automatically. A player relying on it is relying on a safety net, not on his contract.

Rookie deals are where the guarantee has actually moved

The most interesting movement in guarantees in recent years has not been at the top of the veteran market. It has been in the draft.

Drafted players sign four-year contracts under a wage scale, with first-round picks carrying a club option for a fifth year, and the money at each slot is substantially predetermined. What is not predetermined is the guarantee structure, and that is where the negotiation happens. First-round picks receive fully guaranteed four-year deals. Later rounds historically did not, and the boundary between the two has been moving downward.

In 2025 that boundary moved visibly. Second-round picks, who had always signed partially guaranteed deals, began signing fully guaranteed ones, and a group of them held out for weeks to get there. The mechanism was simple and worth understanding, because it is the only leverage a drafted player has. A rookie cannot negotiate his salary, since the scale sets it, so the only variable left is the guarantee, and the only pressure available is refusing to sign while training camp runs without him. Once one club near the top of the round conceded, every player picked behind him had a comparison to point at.

That is collective bargaining without a union at the table, conducted by twenty-two-year-olds with no cards except a signature, and it worked because the slot structure that removes their salary leverage also hands them a perfectly legible precedent. Each pick is worth marginally less than the one above it, so a full guarantee at one selection is an argument for a full guarantee at the next. The scale designed to end holdouts turned out to give rookies the one comparison that makes a holdout coordinated. How the rest of that system is built, slot by slot, is the subject of the rookie wage scale.

The fifth-year option is the other place the guarantee shifted, and it shifted by agreement rather than by attrition. Under the current agreement the option, once exercised, is fully guaranteed rather than guaranteed for injury only, which converts it from a cheap look at a fifth season into a real commitment. Its price is set in tiers according to what the player has actually done: a basic tier, a higher one for players who hit a playing-time threshold across their first three seasons, and higher ones again for players selected to the Pro Bowl on the original ballot, with the top tier priced off the franchise tag figure at the position. A club exercising the option is now making a decision with money attached, which is exactly what the previous version let it avoid.

Why guarantees pile up at one position and disappear at others

Guarantees are not distributed evenly across a roster, and the reason is not that clubs like some players more.

A guarantee is a bet that a player will still be worth his salary in two or three years. The willingness to make that bet tracks two things: how predictable the player's decline is, and how replaceable he is if the prediction is wrong. Quarterbacks score badly on the second and are therefore guaranteed heavily, because a club that has one it believes in has no realistic substitute and knows it. Running backs score well on both, in the sense that the supply is deep and the decline arrives early, and the guarantee market for them reflects that with some brutality.

Positions where physical decline arrives suddenly rather than gradually are also poor candidates for long guarantees, because the club's downside is not a slow slide it can plan around but a cliff it cannot. Positions where the skill is technical rather than athletic hold value longer and can extract more.

This produces a structural unfairness that no amount of individual negotiation fixes. The players taking the most physical punishment, in the roles with the shortest useful careers, are the ones for whom clubs will guarantee the least, precisely because the punishment and the short career are visible in advance. The market prices the risk correctly, and the result is that the people carrying the risk are paid to carry it in instalments they may not last long enough in the sport to collect.

The alternative instrument, for a club that wants one more year without a multi-year guarantee, is the franchise tag: a single fully guaranteed season at a formula price, with no commitment beyond it. Clubs reach for it at exactly the positions where they will not guarantee a long deal, which is not a coincidence.

What a player gives up to get the money guaranteed

Nothing in a negotiation is free, and a guarantee is bought with something.

The most common currency is average annual value. A player who insists on more money guaranteed at signing will often be offered a lower total, because the club is pricing certainty and certainty costs. Announcements do not show this, since the two numbers are reported side by side as though both were achievements, and a deal with a smaller total and a larger real guarantee is frequently the better one.

The second currency is length. Clubs will guarantee more if the contract runs longer, because a longer contract spreads the funding requirement and delays the player's next chance to reprice himself. A player accepting a fifth year in exchange for a bigger guarantee has sold an option he may badly want back if he plays well.

The third is the offset clause, and the fourth is the vesting date, and both get handed over quietly because neither fits in a headline. A player can accept a later vesting date and a full offset and walk away with a guarantee number that looks identical to somebody who conceded neither, and be materially worse off.

The fifth is the structure of the bonus itself. A club can offer a larger guarantee while paying less of it up front, converting bonus into guaranteed salary, which keeps the announced guarantee intact while moving cash from this year into next. That is a real concession by the player dressed as no concession at all.

Which is the honest reason contract analysis in this sport is a specialist activity. Six variables move independently, only two of them are published, and the two that are published are the two the club is happiest to talk about.

Reading a contract announcement without being lied to

The comparison across sports, where a signed number is simply a number the player will receive, is set out separately, and it makes the football case look stranger the longer you look at it. Within this sport, though, five questions will get you most of the way to what a deal is actually worth, and all five are answerable from documents rather than from opinion.

How much is guaranteed at signing, as distinct from total guarantee? This is the single most useful number and it is rarely the one in the headline. Everything else described as guaranteed is conditional on a date.

How much of that is signing bonus? Bonus is cash already moved. Guaranteed base salary is cash that will move if nothing intervenes. They are not the same quality of promise.

What are the vesting dates, and what protection applies before them? Injury-only protection with a vesting date in the third league year is a club option with a nice name.

Is there offset language, and on which years? It changes the value of the guarantee in the exact scenario the guarantee exists for.

What does the third year cost to escape? Add the guaranteed salary that accelerates to the unamortised bonus that accelerates, compare it against the salary saved, and you will know whether the club has a decision or an obligation. Contracts that look identical in the announcement diverge completely at this question.

Answer those and the announcement becomes readable. Skip them and you are quoting a press release with a number in it, which is what most contract coverage is, and which is why the same deals get re-evaluated eighteen months later as though something surprising had happened.

Nothing surprising happened. The paperwork said so at the time.

More on how this sport prices, accounts for and argues about money is collected in the American football archive.