What we know about the typical Polymarket user

X Facebook Threads LinkedIn WhatsApp Mail Add Us On Google By Kaitlyn Radde (Pew Research Center illustration; photos via Getty Images) Prediction markets are an increasingly popular way for people to trade on the outcomes of real-world events. As of April, monthly trading volume on the two largest prediction markets – Polymarket and Kalshi – had reached nearly $24 billion .

To better understand how ordinary people use prediction markets, we collected a sample of nearly 12,000 Polymarket accounts that placed trades on 10 specific high-volume events in early 2026. We then analyzed their publicly available trading activity over the six-week period from May 7 to June 19. Here’s what we found.

This Pew Research Center analysis looks at prediction market trading behavior on Polymarket from May 7 to June 19, 2026.

Pew Research Center does research to help the public, media and decision-makers understand important topics. This research builds on our recent work on gambling and the broader prediction economy – including trading on prediction markets , public attitudes toward sports betting and views on the morality of gambling around the world.

Data on Polymarket users primarily comes from the Polymarket user activity API , which does not include the company’s newer U.S. platform. Kalshi, the other major prediction market platform, does not make this kind of user data available.

We collected trading activity by sampling Polymarket user “wallets” in early May 2026. We first used other Polymarket API endpoints to identify 10 high-volume trading events across a range of topics, collecting the 4,000 most recent trades per event. We used the 16,836 accounts that placed those trades as our sample. We refer to these accounts as “traders” or “users” throughout, though some individuals may run multiple accounts.

We then collected activity for those accounts 15 times between May 7 and June 19, 2026. This process yielded activity for 11,989 active wallets. The API allowed us to collect the 4,000 most recent activities per account per collection using offset pagination. Only 2% of wallets hit this limit, meaning those accounts may have made trades during our analysis period that we did not collect.

To examine the topics that users were interested in, we also recorded the events that these accounts traded on during each data collection. We used Polymarket’s tags to sort events into broad categories, for instance nesting “basketball” under “sports.” Overall, 1% of events fell into multiple categories, while fewer than 1% could not be categorized.