PredictIt Review 2026: The Politics Market That Sat Out the Ban Wave

PredictIt is a real-money market where you buy and sell shares tied to the outcome of political events. That the platform survived the regulatory storm that has redefined the rest of the sector is not an accident, and understanding why is the whole story. We reviewed PredictIt against its own help pages, its terms, the CFTC no-action letters that govern it, the federal court record from its litigation, and a stack of competitor reviews and news coverage. Every legal statement below is dated to August 1, 2026, because in this vertical the facts move week to week.

PredictIt is a real-money market where you buy and sell shares tied to the outcome of political events. Each market is a question — for example, which party will control the Senate after the next election, or whether a specific nominee will be confirmed — and each answer trades as a contract priced between 1 cent and 99 cents. Buy a Yes share at 62 cents and you are effectively paying for the market’s 62 percent implied probability that the event happens. If it happens, that share pays out at $1.00 and you keep the 38-cent gain; if it does not, the share expires worthless. You can also sell before settlement to secure a profit or limit a loss, because shares trade continuously against other users.

What PredictIt is not is a sports betting site or a broad, DCM-listed exchange in the mold of the platforms fighting states in federal court. It does not offer moneylines, spreads or player props. It does not list the sprawling menu of economics, crypto, weather and entertainment contracts you find on the general-purpose venues covered in our prediction markets hub. It is a narrow, politics-first research market that has looked and operated more or less the same way since 2014, and that continuity is central to why it survived the regulatory storm that has redefined the rest of the sector.

The continuous double auction, in plain terms

PredictIt runs an order book, which the platform calls a continuous double auction. Every trade requires a matched counterparty: for you to buy a Yes share, someone has to be willing to sell one at your price, and the reverse for No shares. There is no house taking the other side of your bet the way a bookmaker does, and there are no fixed odds set by the operator. Prices move because buyers and sellers move them. One quirk of the paired-share structure is that Yes and No shares of the same contract are two sides of the same coin: the platform can create a matched pair (one Yes plus one No) for $1.00 total, so a market maker who buys both sides is fully hedged and the book stays balanced. In practice this means liquidity concentrates in the marquee races and can be thin in the long tail, which is a recurring complaint we return to below.

A worked example

Say a market asks which party will control the Senate after the next election, and the Republican Yes share is trading at 55 cents. You think that is too low and buy 100 shares for $55. Three outcomes are possible. If Republicans win, each share settles at $1.00, your 100 shares are worth $100, and your gross profit is $45 — minus the 10 percent profit fee, which takes about $4.50, leaving roughly $40.50 before the separate 5 percent withdrawal fee applies when you cash out. If you change your mind before the election and the price has risen to 70 cents, you can sell your 100 shares for $70, booking a $15 gain that is again subject to the 10 percent fee on the profit. If Republicans lose, the shares settle at zero, you lose the full $55, and no fee is charged on the loss. That single example captures the whole platform: capped stakes, binary settlement at $1 or $0, the ability to trade out early, and a fee that only bites when you win.

Position limits: the feature that defines the whole platform

The single most important design fact about PredictIt is that your money in any one market is capped. Under the original 2014 no-action letter, no trader could hold more than $850 of investment in a single contract, and each market was limited to 5,000 traders. Those caps were modeled on the academic Iowa Electronic Markets and were the price of the CFTC letting the market exist at all. The 2025 settlement raised the per-contract ceiling to a reported $3,500 and removed the 5,000-trader cap. As of August 1, 2026, some markets and some third-party reviews still reference the older $850 figure, so treat the exact live cap as something to confirm inside the specific market you are trading. Either way, the takeaway is the same: PredictIt is built so that no individual can take an outsized position, which keeps stakes small and, crucially, keeps the product from looking like the high-limit sports wagering that has drawn state regulators to its competitors.

PredictIt launched on November 3, 2014, as a nonprofit educational project of Victoria University of Wellington in New Zealand. The academic framing was deliberate. Because online gambling and unlicensed futures trading are heavily restricted in the United States, the university secured a CFTC staff no-action letter in October 2014 that let it run a small-scale, research-oriented market so long as it stayed within tight conditions — the $850 position cap and the 5,000-trader limit chief among them.

The day-to-day operator was Aristotle Inc., the political technology and voter-data company founded by John Aristotle Phillips. The university held nominal academic oversight; Aristotle ran the exchange. The platform grew quickly around the 2016 election, drawing tens of thousands of active traders and building a reputation among journalists, campaign staff and political scientists as a live, money-weighted read on races that polls were struggling to call. By the end of the decade it was handling hundreds of millions of shares a year. Its appeal was never the size of the payouts, which the caps kept modest, but the quality of its signal.

Through the 2016 and 2020 cycles PredictIt became a fixture of political coverage. Reporters cited its prices as a shorthand for where the smart money sat, campaign staff watched its markets for early signals, and academics used its data in forecasting research — exactly the educational use the no-action letter contemplated. It sold itself as a forecasting instrument, and that framing is what let it justify its existence to a federal regulator for the better part of a decade.

The July 2025 settlement and the amended no-action letter

PredictIt prevailed in July 2025. On July 14, 2025, the CFTC issued an amended no-action letter (Letter No. 25-20) that resolved the litigation and let the market keep operating on materially better terms. The agreement did two big things. First, it loosened the caps: the per-contract position limit rose from $850 to a reported $3,500, and the 5,000-trader-per-market ceiling was removed, allowing unlimited participants. Second, it moved leadership out of Aristotle’s hands and into a nonprofit academic structure — reporting describes a consortium of researchers operating the market under an entity styled as the Prediction Market Research Consortium.

This is a genuinely unusual regulatory footing, and it matters for everything downstream. PredictIt does not hold a Designated Contract Market license. Kalshi, ProphetX, Novig and the Aristotle Exchange that Underdog bought are all CFTC-registered exchanges that can self-certify contracts and clear trades through registered clearinghouses. PredictIt instead operates because CFTC staff have formally said, in writing, that they will not recommend enforcement action against it so long as it stays inside the letter’s conditions. It is permission, not a license — narrower, more conditional, and tethered to the small-scale academic character that has always defined the market. For a fuller picture of how federal registration and state gambling law collide across the sector, see our explainer on whether online gambling is legal in the United States.

Here is the part every other PredictIt review buries or misses. Through the most aggressive state crackdown on prediction markets in US history, PredictIt was essentially a bystander. As of August 1, 2026, it has not been named in the cease-and-desist letters, the state lawsuits, or the court orders that have reshaped the rest of the industry. That absence is the story, and understanding why it happened tells you more about PredictIt than any feature list.

The state enforcement wave was aimed at one thing above all: sports. Nevada’s gaming regulators, New York’s attorney general, Ohio’s Casino Control Commission and their counterparts elsewhere went after platforms because those platforms were listing contracts on football games, basketball games and other sporting events that looked, to a state regulator, indistinguishable from a licensed sports bet. A Carson City judge in Nevada described Kalshi’s sports wording as indistinguishable from a Nevada bet; New York’s attorney general framed her case around the fact that Kalshi lets 18-year-olds trade when the state requires 21 for mobile sports betting. Every marquee state action — Nevada, New York, Massachusetts, Ohio, Arizona, Michigan, Washington, Kentucky — has turned on sports event contracts and high-limit, sportsbook-style wagering.

PredictIt offers none of that. It lists political contracts, it caps positions in the hundreds to low thousands of dollars, and it makes no pretense of being a place to bet on the Super Bowl. There is no sports product for a gaming regulator to call unlicensed sports betting, and the position caps make it a poor substitute for a sportsbook even for a determined user. Just as important, the platform’s low volume and explicitly academic, research-oriented framing make it an unappealing target: the money and the political heat are on the multibillion-dollar exchanges, not on a small politics market that has been operating quietly for over a decade.

The result is a rare thing in this sector — a prediction market whose legal position, as of August 1, 2026, is relatively settled. It is not embroiled in the appellate fight that could send sports event contracts to the Supreme Court, it is not facing contempt exposure the way Kalshi is in Nevada, and it is not one of the operators the CFTC has had to defend by suing states. That does not make it invulnerable — a no-action letter is discretionary and can be revisited, as 2022 proved — but it does make PredictIt the calmest corner of an otherwise chaotic map.

One point of genuine confusion deserves its own section, because even careful readers conflate two very different Aristotle entities. Around the time PredictIt won its case, Aristotle also stood up a separate, fully CFTC-registered exchange — the Aristotle Exchange, comprising a Designated Contract Market and a Derivatives Clearing Organization. That was a real license, distinct from PredictIt’s no-action letter, and it was capable of listing and clearing a much broader, higher-limit range of contracts, sports included.

In March 2026, Aristotle sold the Aristotle Exchange DCM and DCO to Underdog, the daily-fantasy-turned-prediction operator, in a deal that gave Underdog its own in-house federal exchange and clearinghouse rather than relying on a third-party partner. The critical detail, and the one most write-ups get wrong, is that Aristotle retained PredictIt. PredictIt was never part of the Aristotle Exchange’s regulatory filings and was not included in the Underdog transaction. Underdog bought the licensed exchange infrastructure; Aristotle kept the politics market that runs on the no-action letter. For what happened next to the buyer — including its own July 2026 agreement to be acquired by IG Group — see our Underdog review. The short version for PredictIt users: the sale changed nothing about how PredictIt operates.

PredictIt’s market menu is the narrowest of any platform we cover, and that is by design. The contracts cluster into a handful of political buckets.

You will not find the economics, crypto, weather, culture and sports contracts that fill out the general-purpose exchanges. If your interest is a single election night, PredictIt is purpose-built for it. If you want a broad event-trading platform, this is not that.

Sports markets: effectively none

Because the brief for any 2026 prediction-market review includes sports, we will be blunt: PredictIt does not offer sports event contracts. There are no moneylines, spreads, totals or player props, and there is no path to bet a game on the platform. This is not an oversight — it is the reason PredictIt stayed clear of the state enforcement wave. Traders who want CFTC-based sports event contracts look to the DCM-licensed exchanges, not to PredictIt.

PredictIt is not free to win on, and its fee structure is aggressive enough that it changes how you should size trades. As of August 1, 2026, two fees apply.

• 10 percent profit fee. PredictIt takes 10 percent of the net profit you make on a position. If you buy a share at 40 cents and it settles at $1.00, your 60-cent gain is taxed 6 cents by the platform. Critically, the fee applies only to profits — there is no fee on a losing trade. But because it is charged per profitable position rather than on your net across the whole site, winners and losers do not cancel out. A day of break-even trading can still cost you money, because you pay 10 percent on every winning position and get nothing back on the losers.
• 5 percent withdrawal fee. When you take money off the platform, PredictIt charges 5 percent of the amount withdrawn. This is separate from and stacks on top of the profit fee, so a dollar you win is hit once when the position profits and again when you cash it out.

Stack those together and the real cost of extracting winnings is meaningfully higher than the headline 10 percent. On a profitable, cashed-out trade you can lose roughly 10 percent of the gain to the profit fee and then 5 percent of the withdrawal to the withdrawal fee. For a platform whose contracts already trade in a tight 1-to-99-cent band, that drag matters: small edges get eaten. This fee model is the opposite of the zero-vig, exchange-style pricing that competitors like Novig market, and it is far heavier than the maker-taker spreads on the big DCM exchanges. If you trade PredictIt, price the fees into every position before you place it.

Before the mechanics, a note that trips up new traders: because fees only touch profits, the smart way to think about PredictIt is in after-fee terms. A share bought at 50 cents does not break even at 50 cents on exit — it has to clear the price you paid plus the 10 percent haircut on whatever gain you book, and then the 5 percent withdrawal fee on the way out. Build that into your entry price and you avoid the most common way casual traders quietly bleed money here.

Funding is straightforward but dated compared with the crypto-friendly newer platforms. PredictIt accepts debit and credit cards and bank transfers for deposits, denominated in US dollars. There is no cryptocurrency funding. Withdrawals are processed back to your linked bank account and carry the 5 percent fee noted above. Payout timelines run on standard bank-transfer speeds rather than the near-instant crypto rails some competitors advertise. Because every withdrawal is taxed at 5 percent, frequent small cash-outs are inefficient; the platform’s economics reward letting a balance ride and withdrawing in larger, less frequent chunks. Confirm current minimums and any card-processing surcharges inside your account, as these can change.

Support is thin and asynchronous. PredictIt does not offer live chat or a phone line. The primary channel is a web contact form on the support section of the site, where you enter a subject and message and a representative replies by email; the platform also fields questions through its social media accounts. For a market that settles binary contracts on hard deadlines, the lack of real-time support is a genuine limitation — if a settlement dispute or a funding issue arises around an election night, you are waiting on an email queue rather than reaching an agent immediately. Response quality is generally reported as adequate for account and settlement questions, but the channel selection is the weakest among the platforms we review.

This is where PredictIt’s unusual legal status cuts against the user. Because PredictIt operates under a federal CFTC no-action letter rather than a state gaming license, it sits outside the state-run consumer protections that cover licensed sportsbooks and casinos. There is no state self-exclusion registry that reaches it, no state gaming-commission complaint process, and no state-supervised recourse to fall back on if something goes wrong. Where a licensed sportsbook must connect problem gamblers to state-mandated resources and honor a statewide self-exclusion list, PredictIt has no such obligation and no such backstop.

The minimum age is 18, not the 21 that most states require for mobile sports betting — a gap that has been at the center of state complaints against other prediction markets and that applies here too, even though no state has moved against PredictIt over it. On the mitigating side, the platform’s structural caps do a lot of quiet work: with positions limited to the hundreds or low thousands of dollars per contract and no sports product to chase, the scale of potential loss is far smaller than on a high-limit sportsbook or a broad exchange. PredictIt does not, as of August 1, 2026, ship the voluntary deposit limits, deposit alerts and self-exclusion tools that the most consumer-protective new entrant in the sector, FanDuel Predicts, has built. If you need formal responsible-gambling infrastructure, understand that on PredictIt the guardrails are mostly the position caps, not a dedicated protection program.

Tax treatment of prediction-market winnings remains unsettled across the entire sector, and PredictIt is no exception. The IRS has published no classification guidance of any kind on event-contract income, and the industry norm has been not to issue 1099-B forms the way a traditional brokerage would for securities. Whether your PredictIt gains are best treated as gambling income, ordinary income or something else is a question the law has not cleanly answered, and it can differ from how a brokerage-distributed contract is reported. You are responsible for tracking your own results and reporting them, and given the platform’s fee structure, keeping clean records of cost basis, profit fees and withdrawal fees is worth the effort. Consult a tax professional; do not assume the platform will hand you a tidy tax form.

Market integrity has become one of the defining worries of the prediction-market boom, and it is worth measuring PredictIt against that backdrop. Across the broader sector in 2025 and 2026, regulators and prosecutors brought a string of insider-trading cases: the Department of Justice charged an Army master sergeant over trades tied to a classified military operation, the CFTC charged a Google engineer over trades on year-end search-trend contracts, and a White House staffer became the subject of a federal investigation over trades on a rival platform. The common thread was people trading contracts on outcomes they had nonpublic information about.

PredictIt’s structure blunts several of these risks by design. The politics-only menu means there are no player-injury, referee or game-outcome contracts of the kind sports leagues have flagged as easily manipulated. The position caps put a hard ceiling on how much anyone can profit from an information edge in a single market — an insider who knows an outcome still cannot deploy more than the per-contract limit — which caps both the incentive and the damage. And the market has, at times, restricted trading by people directly connected to the outcomes it lists. None of this makes PredictIt immune to someone trading on inside political knowledge, and its liquidity is thin enough that a determined actor could move a small market. But the same caps that frustrate large traders also make PredictIt a structurally poor venue for large-scale manipulation, which is one more reason it has stayed off regulators’ enforcement radar.

To understand PredictIt’s position you have to understand the fight it is not part of. Across the sector, the core legal question is whether event contracts are federally regulated swaps under the Commodity Exchange Act, which would vest exclusive jurisdiction in the CFTC and set aside state gambling law, or unlicensed sports betting subject to every state gaming regulator. Federal appellate courts have split on that question for the DCM-licensed exchanges, and the dispute is widely expected to reach the Supreme Court. The CFTC under its 2025-2026 leadership has taken the platforms’ side aggressively, even suing states that moved against federally registered operators.

PredictIt sits to the side of all of it. It is not a Designated Contract Market, so it is not the subject of the DCM preemption cases; it does not list sports, so it is not what the states are angry about; and it operates on an individually negotiated no-action letter that a federal court fight already tested and upheld. In a sector where a single appellate ruling can rewrite every platform’s legal status overnight, PredictIt’s exposure to that volatility is unusually low. The flip side, worth repeating, is that a no-action letter is discretionary staff forbearance, not a statutory right — the 2022 withdrawal proved the CFTC can move against it — so its calm depends on the platform continuing to stay inside the narrow, small-scale lane its letter defines.

PredictIt is a poor fit for the mass-market bettor and an excellent fit for a specific user. It suits the political obsessive who wants to put money behind a read on a race, the forecaster or analyst who values a live probability more than a payout, the researcher pulling market data, and the small-stakes trader who is comfortable with caps and does not care about sports. It is a poor fit for anyone who wants high limits, a broad menu across sports and economics, a modern mobile app, zero-vig pricing, or a promotional bonus to get started. If you recognize yourself in the first group, PredictIt has no real substitute; if you are in the second, almost any other platform we cover will serve you better.

PredictIt occupies a category of one, but it is worth situating against the platforms readers most often weigh it against.

Across user reviews and news coverage, the recurring gripes about PredictIt are consistent.

• The fees. The 10 percent profit fee plus the 5 percent withdrawal fee is the single most cited frustration. Because the profit fee is charged per winning position rather than on net results, active traders feel it acutely, and it can turn a marginally profitable strategy into a losing one.
• Position caps. Even at the raised limits, serious traders find the per-contract ceiling too low to deploy meaningful capital, which is exactly the constraint the no-action letter is built around. What protects the platform legally frustrates its heaviest users.
• Thin liquidity in the long tail. Marquee races are liquid; obscure or early-cycle markets can have wide spreads and little depth, making it hard to enter or exit at a fair price.
• No app and slow support. The web-only experience and the email-only support channel feel dated next to app-first competitors, especially during high-stakes election nights when markets move fast.
• Politics-only scope. Users who want to trade sports, economics or crypto have to go elsewhere; PredictIt does one thing.

PredictIt is a specialist tool, and judged as one it is very good. If you follow American politics closely and want a live, money-weighted market on elections, nominations and control of Congress, nothing else has PredictIt’s focus, its track record or its settled legal footing. The 2025 settlement gave it higher limits and a durable, court-tested basis to keep operating, and its politics-only, low-cap structure is precisely what kept it out of the enforcement wave that has swallowed the rest of the sector. That stability is a real, underrated asset in a market where competitors are fighting for survival in federal appeals courts.

The costs are just as real. The stacked 10 percent profit and 5 percent withdrawal fees are heavy and asymmetric, the position caps frustrate anyone with capital to deploy, there is no app, support is email-only, and the scope is narrow to the point of exclusion — no sports, no economics, no crypto. This is not the platform for a trader who wants breadth, liquidity or a modern app experience; those users belong on a DCM-licensed exchange. But for the political forecaster who values a focused market with a decade-plus of history and, as of August 1, 2026, the calmest legal position in the business, PredictIt still earns its place. Price in the fees, size trades against the caps, and treat it for what it is: the politics market that sat out the ban wave and is still standing.

This review draws on PredictIt’s own help, support and terms pages; the CFTC no-action letters governing the platform, including the amended letter issued July 14, 2025; the federal court record from its 2022-2025 litigation; contemporaneous news coverage of the settlement and of the March 2026 Aristotle Exchange sale to Underdog; and multiple competitor and app-store reviews, which we checked for the common error of overstating prediction-market legality. Where sources conflicted — most notably on the current per-contract position limit — we flagged the uncertainty rather than picking a number. Every legal statement is dated to August 1, 2026, and reflects a fast-moving regulatory landscape that readers should re-verify before relying on it.