State Taxes on Prediction Markets in 2026: Platform Levies and the Trap States

2026 is the year states stopped ignoring prediction markets. North Carolina, Kentucky, and Illinois have all passed taxes aimed at Kalshi and Polymarket, and the CFTC is suing over every single one of them. None of that reaches your personal tax bill directly, but it changes the regulatory ground you're trading on, and a separate, older set of state rules already does reach you directly depending on which federal tax treatment you pick.

The three state platform taxes

These are levies on the operators, Kalshi and Polymarket themselves, not on individual traders. They matter to you as a signal of how hostile a given state is to these platforms, and because litigation outcomes here could eventually reshape how the whole market is regulated.

New Jersey legislators have floated a roughly 30% exchange tax, and Iowa's SF 2470 proposed a $10 million permit fee plus a tax; neither has passed as of this writing. Every one of these is a live, moving target, and dates and rates here should be treated as a snapshot, not a permanent fact.

The CFTC is suing all of them

By late June 2026, the CFTC had sued at least nine states over prediction-market restrictions and taxes, including Arizona, Connecticut, Illinois, Kentucky, Nevada, and New Jersey. The core argument surfaced most clearly in Nadex v. Nevada, now before the Ninth Circuit: the CFTC filed an amicus brief on February 17, 2026 asserting "exclusive jurisdiction over the U.S. commodity derivatives markets, including event contract markets." States are calling event contracts gambling; the platforms and the CFTC are calling them swaps regulated under the Commodity Exchange Act. A Ninth Circuit or eventual Supreme Court ruling on that jurisdictional question could unwind every state platform tax above, or entrench them, and nothing is settled as of mid-2026.

None of this taxes you directly, so what does

The platform taxes above are levied on Kalshi and Polymarket's revenue, not on your winnings. Your actual state exposure comes from something older and less headline-grabbing: how your state taxes the income your federal return already characterizes. If your state starts from federal AGI, and most states with a broad income tax do, then whichever treatment you pick on your federal return (covered in our overview post) flows straight through to your state bill too, cap and all. That includes the federal OBBBA 90% gambling-loss cap: because state taxable income typically starts from federal AGI, a state that conforms to federal rules inherits that cap automatically.

The trap states: taxed on gross winnings, no loss deduction

A handful of states go further and simply don't allow a gambling-loss deduction at the state level, regardless of what you can deduct federally. If you use gambling treatment in one of these states, you're taxed on your gross winnings with no offset for what you lost, on top of whatever the federal 90% cap already does to you. Three are confirmed by primary state sources:

• Connecticut. Gambling losses aren't deductible for state income tax on residents (non-resident winnings, separately, aren't taxed by Connecticut at all).
• Massachusetts. The state doesn't adopt IRC §165(d); it only allows a narrow deduction for losses at Massachusetts-licensed gambling establishments, so out-of-state or online losses, which is what prediction-market trading is, aren't deductible.
• Rhode Island. Gambling losses aren't deductible; federal AGI flows onto the state return with no line to subtract them.

North Carolina and Illinois are also cited, in a secondary practitioner source rather than either state's own tax authority, as disallowing the gambling-loss deduction too. Treat that one as plausible but unconfirmed; check the current position with a professional or the state's own guidance before relying on it, especially given how fast both states' prediction-market rules are moving in 2026.

Rates, for context

Top marginal state individual income-tax rates for 2026, from the Tax Foundation: California 13.3%, New York 10.9%, New Jersey 10.75%, Hawaii 11%, Rhode Island 5.99% (a trap state, see above), Connecticut 5.0% (also a trap state), Massachusetts 5% (9% on long-term gains, also a trap state), Illinois 4.95% flat, North Carolina 4.5% flat, Kentucky 4.0% flat. Texas, Florida, Nevada, Wyoming, Washington, and South Dakota have no individual income tax at all, so trading gains go untaxed at the state level regardless of federal treatment.

Why this matters when you pick a method

The gambling method is already usually the worst of the four federal treatments, and the 2026 OBBBA cap (covered in our deep dive on the cap) made it worse. If you're a resident of Connecticut, Massachusetts, or Rhode Island, gambling treatment gets worse again at the state level, taxed on gross winnings twice over, federal and state, with no loss offset either time. That's exactly the kind of gap a four-method comparison is built to surface before you file, not after.