Kalshi saw over $1.8 billion in trading volume on March Madness, more than three times what it took on the 2024 presidential election.
State-regulated sportsbooks are estimated to have generated over $3 billion in handle on the tournament. Handle, or total dollars wagered, differs from prediction market trading volume, which counts every buy and sell side of a contract, even those traded multiple times. As such, they are not apples-to-apples, and the gap between them is difficult to quantify.
Kalshi trading volume is tracked independently by Legal Sports Report, which reviewed activity on all 134 men’s and women’s NCAA tournament games in addition to prop markets and parlays associated with each contest.
March Madness betting sets record
The men’s national championship game between UConn and Michigan saw $105.5 million in trading volume, the most of on a single basketball game in Kalshi history, a record it broke repeatedly throughout the tournament. It is also the most Kalshi has taken on a single sporting event after Super Bowl LX.
The women’s tournament accounted for $177 million in trading volume, with $16 million on the championship game between UCLA and South Carolina.
For the full tournament, 73% of total volume was placed live, during games, compared to an estimated 55% to 60% at traditional U.S. sportsbooks, according to Citizens, which tracked in-play activity across every round.
Kalshi data comes with caveats
The championship number carries an important caveat. Because Kalshi’s game-winner markets for championship games operate as a season-long futures market rather than a standalone game line, it encompassed 116 possible outcomes.
LSR’s $105.5 million figure represents only volume placed on either UConn or Michigan to win the national title between April 4th and April 6th, in addition to other game-specific markets placed in that time span, such as spread, total, first half winner and same-game parlays.
The same challenge applies when comparing Kalshi’s game-level figures to its Super Bowl data. Between the NFL conference championship games and Super Bowl LX, $214 million in trading volume was placed on either Seattle or New England to win, with an additional $155 million traded on props. The $358 million figure listed on Kalshi’s website includes betting on all 32 teams from the start of the season.
They are offered here to illustrate how difficult it remains to benchmark Kalshi’s event-level volume against itself, let alone against sportsbook handle data.
Sports pushes Kalshi past monthly record
March was Kalshi’s largest month on record for trading volume, reaching $13 billion, up more than 25 times year-over-year and surpassing its previous all-time monthly high set in February.
Sports markets accounted for 86% of March’s total volume. Parlays accounted for roughly 9% of all platform volume for the month.
Kalshi first launched sports trading in January 2025, three days after the inauguration of President Donald Trump, whose son serves as a strategic advisor to the company.
Since then, it has taken over $60 billion in total trading volume, 86% of which has been on sports event contracts. The rapid ascent came after years in which the platform was focused primarily on financial and political event contracts.
Kalshi beats sportsbooks on (some) pricing
Kalshi offered the best pricing on moneyline and game-total markets in a leading 41 of 67 men’s tournament games, compared to major sportsbooks, according to Citizens.
Citizens calculated the following average implied vig across the full tournament:
- Kalshi: 4.13%
- DraftKings: 4.35%
- FanDuel: 4.45%
- Fanatics: 4.48%
The gap may be a fraction of a penny on the dollar, but it translates to Kalshi retaining approximately 7% less of every dollar wagered than its competitors on those markets, according to Citizens.
DraftKings’ and FanDuel’s own prediction market platforms were tracked separately and both were found to offer inferior pricing compared to their sportsbook products.
Sportsbooks still own key markets
Kalshi charges a transaction fee on every contract traded rather than building margin into its prices through traditional vig, but structures those fees differently depending on who is trading. Average bettors pay the standard retail fee while high-volume traders are incentivized to provide liquidity that tightens markets under more favorable terms. Citizens noted that this structure “appears to favor a specific cohort of sophisticated users, many of whom may otherwise wager on offshore platforms.”
But the pricing edge comes with a substantial asterisk. The Citizens analysis covers but a slice of the overall betting menu. Spreads, player props, and parlays, which collectively represent the majority of handle at traditional sportsbooks, were not included.
On those markets, Citizens does not find Kalshi to be the pricing leader, and the firm maintained its view that prediction market platforms pose “limited impact” to sportsbook profitability in regulated states. Kalshi’s parlay product, meanwhile, is still developing. Both DraftKings and FanDuel have announced plans to invest hundreds of millions of dollars into building out their prediction market parlay offerings.
Sports betting vs. financial trading dispute wagers on
More than 20 states have filed suit against Kalshi and other prediction market platforms, arguing that their sports offerings are not financial derivatives subject to exclusive federal jurisdiction, but are instead sports betting products that states have the right to regulate under their own laws.
Earlier this week, a New Jersey appellate court ruled 2-1 in Kalshi’s favor, while last week the CFTC filed suit against Arizona, Connecticut and Illinois, in an attempt to asserting its federal authority over the fast-growing industry.
The agency has pledged forthcoming “future proof” rules governing sports events contracts and is gathering public input with over 700 comments so far, including from the NCAA which has repeatedly called to apply sportsbook-style integrity guardians to sport prediction markets.
Data Analyst Eric Ramsey contributed to this report.
