Tax Center
Are Prediction Markets Taxed as Gambling?
By Owen Monagan ·
There is no IRS ruling, notice, or regulation treating prediction-market event contracts as gambling, and the regulatory record points the other way: Kalshi and Polymarket US are CFTC-designated contract markets, and a federal appeals court has held their sports contracts are likely swaps under federal commodities law. But how they are taxed is a separate and genuinely unsettled question. This is information, not tax advice.
This guide covers what the law currently says about whether these are wagers, why that does not by itself settle the tax question, what gambling treatment would actually cost you in 2026, and how the two trading treatments compare. Federal only; state treatment is out of scope.
What does the law currently say about whether these are gambling?
On the regulatory question, there is a clear current answer, and it is not gambling.
Kalshi and Polymarket US operate as designated contract markets regulated by the Commodity Futures Trading Commission, not as state-licensed sportsbooks. Two consequences are visible to any trader:
- You can trade sports event contracts in Texas and California. Sports betting is illegal in both. That is roughly 70 million people between them, and they are two of the ten states with no legal sports betting at all. The contracts are available anyway because federal commodities regulation is asserted to preempt state gambling law.
- Kalshi's minimum age is 18. Most states set sportsbooks at 21. A CFTC-regulated derivatives exchange follows the general age of majority, because it is regulated as a financial venue rather than a gambling one.
In April 2026 the U.S. Court of Appeals for the Third Circuit addressed this directly. In a 2-1 decision (KalshiEX, LLC v. Platkin, No. 25-1922), the court affirmed a preliminary injunction barring New Jersey from enforcing its gambling laws against Kalshi, holding that Kalshi was likely to succeed in showing that sports event contracts are "swaps" under the Commodity Exchange Act and that the CEA preempts state gambling law as applied to them. Judge Roth dissented, writing that the contracts are virtually indistinguishable from bets offered by state-regulated sportsbooks.
Two honest caveats on that ruling. It affirmed a preliminary injunction, a likelihood-of-success standard, not a final merits judgment. And it is a regulatory holding, not a tax one; it says nothing about §165(d), §1256, or how you file.
The fight is also live. Minnesota enacted an outright ban, Illinois enacted a first-in-the-nation per-contract tax on prediction markets, and the CFTC has brought legal action against nine states asserting federal preemption. Sports contracts are currently restricted or unavailable in roughly eight states as a result of state enforcement, including Nevada under a court order. Those are contested state actions against a controlling federal position, not settled law. But the availability map genuinely does vary by state, and it moves.
Does that settle how they are taxed?
No. Regulatory characterization and tax characterization are different questions decided by different bodies, and the IRS has answered neither.
What the regulatory record does is make the wagering argument harder to sustain. IRC §165(d) limits deductions for "losses from wagering transactions." A contract traded on a CFTC-designated exchange, with a counterparty on the other side rather than a house, is a difficult fit for that phrase, but "difficult fit" is not "resolved," and no ruling has resolved it.
Note the direction the swap argument cuts. A swap is a financial instrument, which is an argument against wagering treatment. It is also an argument against §1256, for the reasons below. The two points are not in tension; they simply push toward the middle of the spectrum.
What would gambling treatment actually cost you?
Two things bite, and the second one is worse.
The 90% cap. For tax years beginning after December 31, 2025, §165(d) as amended by the One Big Beautiful Bill Act allows a deduction equal to 90% of wagering losses, still only to the extent of wagering gains. The disallowed 10% is taxed as income you never kept.
Work it through on a break-even year. Say you trade daily, win $100 on your winning days and lose $100 on your losing days, and finish exactly flat: $36,500 of winnings against $36,500 of losses.
| Step | Amount |
|---|---|
| Winnings reported | $36,500 |
| Losses incurred | $36,500 |
| Deduction allowed (90%) | $32,850 |
| Taxable phantom income | $3,650 on a break-even year |
The itemizing requirement. Per IRS Topic No. 419, "you may deduct gambling losses only if you itemize your deductions on Schedule A." Most filers take the standard deduction. For that trader, the deduction for those losses is zero, and the same break-even year is taxed on the full $36,500.
Our guide on the OBBBA 90% gambling-loss limit covers that rule and its repeal prospects in detail.
How do the two trading treatments compare?
Ordinary capital treatment. Gains and losses are capital. Anything held under a year is short-term and taxed at ordinary rates, which covers nearly every prediction-market position. Losses net against capital gains in full. Under §1211(b), net losses beyond that deduct up to $3,000 per year against ordinary income ($1,500 if married filing separately). At a 22% federal rate, that $3,000 is roughly $660. What remains carries forward under §1212(b); the IRS states no expiration for the carryover, so unused losses keep offsetting future gains and income until they are used up.
Section 1256 treatment. Under §1256, gains and losses are treated as 60% long-term and 40% short-term regardless of holding period, with mark-to-market at year end, reported on Form 6781. A net §1256 loss can also be carried back three years against prior-year net §1256 gains under §1212(c), though it cannot create or increase a net operating loss. Better rates, plus a refund path a pure carryforward does not offer. See our Section 1256 60/40 guide for the mechanics.
It is also the aggressive end of the spectrum, and the argument runs both ways:
- For: these exchanges are CFTC-designated contract markets, which satisfies the "qualified board or exchange" prong at §1256(g)(7)(B): a domestic board of trade designated as a contract market by the CFTC. A binary contract with a fixed payout can be read as a nonequity option.
- Against: §1256(b)(2)(B) excludes swaps and "similar agreements" from the definition, and event contracts are swaps under the Commodity Exchange Act, the characterization the Third Circuit reinforced in April 2026. On top of that, the IRS has never ruled that event contracts qualify, and mark-to-market cuts both ways: it can accelerate tax on open winners you have not sold.
Cost versus fit: the three treatments side by side
Cost and accuracy are different questions, so it helps to rate them separately.
| Treatment | What it costs you | How well the argument holds up |
|---|---|---|
| Gambling (§165(d)) | Highest. 90% cap on loss deductions, itemizing required, phantom income on a break-even year | Weakest. Sits against the regulatory basis the venues operate on, though the IRS has never ruled it out |
| Ordinary capital | Moderate. Losses net against capital gains, then $3,000/yr against income, carried forward | Strongest. Plain capital treatment of a financial contract, with no special regime to qualify for |
| Section 1256 (60/40) | Lowest. Blended 60/40 rates, plus a three-year loss carryback | Aggressive. Good designated-contract-market argument, but untested and exposed to the swap exclusion |
Note: the §1256 row assumes prediction-market event contracts qualify as §1256 contracts, which is not settled. The "argument" column is our reading of the current record, not a professional opinion and not a recommendation. Which treatment fits your facts is a decision for you and a qualified tax professional.
Is filing as gambling the "safe" choice?
Plenty of traders reason this way: this feels like gambling, so I will file it as gambling, pay more, and stay out of trouble.
That instinct is worth examining rather than acting on. Paying more tax is not the same as filing correctly. The IRS does not give credit for volunteering extra tax under a characterization that does not match the instrument, and how an activity feels has no bearing on what it legally is. The mechanism here is a regulated exchange and a financial contract: you are not betting against a house, you are buying and selling a contract with someone on the other side, at a price that moves.
The honest summary is that gambling treatment is simultaneously the most expensive of the three and the hardest to square with the regulatory record, so it is not obviously the cautious option. That is not a recommendation to file any particular way. It is a reason to make the decision deliberately, with a professional, rather than defaulting to the most costly treatment on the assumption that costly means safe.
Run your own numbers
The spread between these treatments depends on how much you traded, not just how you finished. Because §165(d) works on gross winnings and losses while capital treatment nets, a high-churn break-even year and a low-churn break-even year look identical on a P&L and completely different on a return.
Our free prediction market tax calculator prices all three treatments for a year of trading, with no sign-up and no email. The itemize toggle is the one worth playing with, since it is the difference between a partial deduction and none at all. The figures are estimates under stated assumptions, not a filed return.
What is genuinely unsettled
To be explicit about the limits of everything above:
- There is no IRS guidance on event contracts. No ruling, no notice, no regulation. Whether they are §1256 contracts is a real open question, and whether they are wagering under §165(d) is technically open too, though the designated-contract-market status is a strong argument that they are not.
- Reclassification can reach back. If the IRS issues guidance or wins an audit challenge, the effect is not necessarily prospective only. It can reach open prior years, with back tax and interest on returns already filed. That risk is largest for the most aggressive position.
- The state fight is live and moving. Bans, taxes, and preemption litigation are all in progress. That affects where you can trade, not how you file federally.
- This is federal only. State tax treatment varies and is not covered here.
For the bigger picture, start at the Prediction Market Tax Center, and see our related guides on whether Kalshi winnings are taxable and what to track for taxes.
The bottom line
No IRS guidance characterizes prediction-market contracts as gambling, and the current regulatory record (CFTC-designated exchanges, a federal appeals court holding these are likely swaps, and availability in states where sports betting is illegal) points away from it. But the tax question is separate from the regulatory one and remains genuinely open, with §1256 sitting at the aggressive end of the spectrum and ordinary capital treatment in the middle. Keep complete records of every trade, understand what each treatment would cost you, and decide which one fits your facts with a qualified tax professional before you file.
Status as of August 6, 2026. The litigation and legislation described here are active and may change; verify the current position before relying on it.