5 Best Prediction Markets for US Traders (Legal Options)

If you've tried to trade on a prediction market from the US, you've probably hit a wall. Polymarket blocked US IPs in 2022. Betfair hasn't accepted American customers since 2010. And most crypto-based platforms either geo-fence Americans or pretend not to notice them — until they do.

But the picture isn't all restrictions. The Commodity Futures Trading Commission (CFTC) started approving event contracts in 2020, and the market has moved fast since then. Kalshi launched as the first fully regulated prediction market exchange in the US. PredictIt still operates under a no-action letter. And academic platforms like the Iowa Electronic Markets have been running legally since 1988.

The key distinction: the CFTC treats prediction markets as derivatives. If a platform wants to offer contracts for real money, it needs to register as a Designated Contract Market (DCM) or obtain a specific exemption. Platforms that skip this step are operating in a gray area at best.

State law adds another layer. While the CFTC provides the federal framework, some states have their own restrictions on event contracts, particularly those that look like gambling. New York, for instance, has been more aggressive than most in scrutinizing these products.

Kalshi is the cleanest legal option for US traders. Full stop. The company registered as a Designated Contract Market with the CFTC in 2020 and launched publicly in mid-2021. Every contract on the platform has been reviewed and approved by regulators before it goes live.

That regulatory status comes with real tradeoffs. Kalshi can't just list any market it wants. In 2023, the CFTC initially blocked Kalshi's Congressional control contracts before a federal court overruled the decision. The legal battle took over a year and went all the way to the DC Circuit Court of Appeals. As of 2025, Kalshi offers political contracts, economic data contracts (CPI, jobs numbers, GDP), weather events, and some entertainment markets.

The trading experience itself is solid. You deposit USD via ACH or wire transfer. Contracts trade between $0.01 and $0.99, where the price reflects the market's implied probability. If you buy a "Yes" contract at $0.65 and the event happens, you get $1.00 — a profit of $0.35 per contract. If it doesn't happen, you lose your $0.65.

Fees are reasonable: $0.01 per contract per side for standard markets, with no withdrawal fees for ACH. There's no cap on position size for most contracts, though liquidity varies. Major political markets might have millions in open interest. A niche weather contract might have a few thousand dollars.

KYC is mandatory — you'll submit a government-issued ID and SSN during signup. Withdrawals typically take 1-3 business days via ACH.

Bottom line: If you want to trade prediction markets with real money in the US without worrying about legal gray areas, Kalshi is the default choice. The market selection is narrower than offshore platforms, but you're trading on the right side of the law.

PredictIt has been the go-to political prediction market for US traders since 2014. It operates under a CFTC no-action letter granted to Victoria University of Wellington, which means the CFTC agreed not to take enforcement action against the platform — as long as it follows certain rules.

Those rules create limitations. Each trader can invest a maximum of $850 per contract. The platform can list no more than 5,000 traders per market. And every market must serve an academic research purpose. In practice, that last requirement is loosely enforced — PredictIt lists markets on everything from election outcomes to Fed rate decisions.

The bigger issue: the CFTC moved to revoke PredictIt's no-action letter in late 2022. A federal judge in Texas blocked that revocation, and as of early 2026, PredictIt continues to operate. But the legal situation is precarious. The no-action letter could be terminated with a few months' notice if the CFTC decides to push again.

Trading mechanics are straightforward. You fund your account with a credit card, debit card, or bank transfer. Contracts trade as shares priced between $0.01 and $1.00. PredictIt takes a 10% fee on profits and a 5% fee on withdrawals — significantly higher than Kalshi's fee structure.

Despite the fees and position limits, PredictIt has a loyal user base. Its political markets are among the most actively traded prediction markets in the US, particularly during election cycles. The $850 cap per contract keeps it from being a serious trading venue for large positions, but it's perfectly adequate for retail-scale trading.

Bottom line: PredictIt works, but the legal foundation is shakier than Kalshi's and the fees are steep. If the CFTC no-action letter survives, it's a solid option for political junkies who want skin in the game.

The Iowa Electronic Markets (IEM) have been running since 1988, making them the longest-operating prediction market in the US. Run by the University of Iowa's Tippie College of Business, the IEM operates under a CFTC no-action letter with stricter limits than PredictIt: a maximum investment of $500 per trader.

Market selection is extremely limited. The IEM typically runs markets on presidential elections, congressional elections, and occasionally Federal Reserve decisions. Don't expect markets on the Oscars or hurricane landfalls — this is an academic research platform first.

Registration requires a university-affiliated account, though non-students and non-faculty can still register. The interface looks like it was built in 2003 — because it was. Expect a barebones trading experience with no mobile optimization, no real-time charts, and manual order entry.

So why mention it? Track record. The IEM has outperformed major pollsters in predicting presidential election outcomes in 9 out of the last 11 election cycles. That record has produced dozens of published academic papers on information aggregation and market efficiency. If you're interested in prediction markets as a forecasting tool rather than a trading venue, the IEM's historical data is a goldmine.

Bottom line: The IEM is more of a research artifact than a trading platform. But it's fully legal, has a decades-long track record, and the $500 buy-in is low enough to dip your toes in.

Metaculus isn't a prediction market in the financial sense. No money changes hands. You don't buy contracts. Instead, you submit probability estimates on questions — and you're scored based on accuracy over time.

That distinction matters legally: because there's no financial component, Metaculus operates without any CFTC restrictions. Anyone in any country can participate.

The question catalog is worth your attention. Metaculus covers AI timelines, geopolitical events, scientific breakthroughs, climate milestones, pandemic trajectories, and economic indicators. As of early 2026, the platform has resolved over 10,000 questions with a combined 3 million+ predictions.

What makes Metaculus worth your time is the community's accuracy. The platform's aggregate predictions have been consistently well-calibrated — when Metaculus says something has a 70% chance, it happens roughly 70% of the time. That's better than most individual experts, most polls, and most pundits.

The scoring system uses a log-based proper scoring rule, which means you're rewarded for honest, well-calibrated estimates and penalized for overconfident or underconfident guesses. Top forecasters build reputations that carry weight in policy circles — several Metaculus forecasters have been cited in congressional testimony and consulted by government agencies.

Bottom line: You won't make money on Metaculus, but you'll become a better thinker. It's also the single best free tool for checking what calibrated, thoughtful people believe about future events.

Manifold uses a play-money currency called Mana (M$). You start with M$1,000 and can earn more through accurate predictions, referrals, and daily login bonuses. Mana can't be converted to cash, but it can be donated to partner charities — so there's a real-world incentive to forecast well.

The market creation model sets Manifold apart. Anyone can create a market on anything. Want to bet on whether your city council will approve a new bike lane? Make a market. Curious if GPT-5 will launch before July? Someone's probably already made that market. As of 2026, Manifold has hosted over 500,000 user-created markets.

This openness creates both breadth and noise. Some Manifold markets are well-researched with hundreds of participants. Others are inside jokes with three traders. The platform's scoring and reputation system helps surface quality, but you'll need to learn which markets are worth trusting and which are just for fun.

Because no real money is involved, Manifold faces zero regulatory restrictions in the US. The charity donation angle — Manifold has facilitated over $1 million in charitable giving through its platform — adds a warm-fuzzy layer that pure play-money platforms lack.

The interface is modern, mobile-friendly, and updated frequently. Market resolution is handled by the market creator, which occasionally leads to disputes. Manifold has a community mediation process for contested resolutions.

Bottom line: Manifold is the best free prediction market for sheer variety. You won't profit financially, but you'll sharpen your forecasting skills and maybe direct some charity dollars along the way.

The elephant in the room. Polymarket is the world's largest prediction market by volume — it processed over $3 billion in trades during the 2024 US presidential election cycle. The interface is polished, the liquidity is deep, and the market selection is broad.

But Polymarket doesn't accept US traders. Not officially. The platform settled with the CFTC in January 2022 for $1.4 million after operating an unregistered exchange. As part of that settlement, Polymarket agreed to block US customers and not offer services to Americans.

Some US users access Polymarket via VPN. This violates Polymarket's Terms of Service and arguably violates CFTC regulations. The practical enforcement risk for individual retail users is low — the CFTC has historically focused on platforms, not individual traders. But "low risk" isn't "no risk." If you use a VPN to access Polymarket and withdraw profits, you're creating a paper trail that could theoretically attract attention.

There's also a practical risk: if Polymarket ever does a KYC sweep or tightens its geo-fencing, US-based accounts could be frozen. The platform runs on the Polygon blockchain using USDC, and while smart contracts are permissionless, the front-end is not.

We're not telling you what to do. We're telling you what the rules say. CFTC regulations apply to US persons regardless of which VPN server they're connected to.

DraftKings, FanDuel, and BetMGM all offer proposition bets that look and feel like prediction markets. "Will Taylor Swift perform at the Super Bowl halftime show?" is functionally identical to a prediction market contract.

These platforms are legal in 30+ states (as of early 2026) because they're regulated as sports betting operators under state gaming commissions — not as derivatives exchanges under the CFTC. Different regulatory box, same basic mechanic.

The catch: prop bet markets are typically less efficient than dedicated prediction markets. The vig (the house edge built into the odds) is usually 5-10%, compared to Kalshi's 1-2% fee. You're paying a premium for the convenience of using a mainstream app.

Also, sports betting apps limit the types of propositions they'll offer. Political markets are generally off-limits due to state gaming regulations. Economic data markets don't exist. You're restricted to sports, entertainment, and cultural events — and even those categories are constrained by what each state's gaming commission approves.

If you already have a DraftKings or FanDuel account and just want to make occasional predictions, prop bets work fine. But for serious prediction market trading, they're an expensive substitute.

Prediction market profits are taxable. The IRS hasn't issued specific guidance on prediction market taxation, which means the rules depend on how the contracts are classified.

Kalshi contracts are traded on a CFTC-regulated exchange, and the company issues 1099 forms. Many tax advisors classify Kalshi profits under Section 1256, which provides a blended rate: 60% of gains are taxed as long-term capital gains and 40% as short-term, regardless of holding period. For someone in the 35% federal bracket, this blended rate works out to about 26.8% — significantly better than the 35% ordinary income rate.

PredictIt is murkier. The platform issues 1099-MISC forms, and many tax professionals treat PredictIt income as ordinary income (taxed at your marginal rate). Some traders have argued for Section 1256 treatment, but there's no IRS ruling confirming that interpretation.

Play-money platforms like Metaculus and Manifold — no taxable event. You're not receiving income. Manifold's charity donations may be deductible if the receiving organization is a 501(c)(3), but that's between you and your accountant.

The bottom line: keep records of every trade, every deposit, and every withdrawal. Use a tax professional who understands derivatives. Don't assume that because a platform doesn't issue a 1099, you don't owe taxes — the IRS expects you to report income regardless of whether you receive a form.