Are UFC Fights Getting Longer? What 1,735 Bouts Show

Quick answer

Yes, compared with the early 2000s. But the rise appears to have flattened. A 2026 study compared 1,735 men's UFC bouts across three two-year windows. Mean duration rose from 495.75 seconds in 2003-04 to 636.53 seconds in 2013-14. It reached 650.14 seconds in 2023-24. The last two periods weren't statistically different. Modern UFC fights are longer than their early-2000s counterparts. The evidence doesn't show that they keep getting longer today.

Data snapshot: 1,735 decided men's UFC bouts from 175 events in 2003-04, 2013-14 and 2023-24; plus a 1,660-bout three-round sensitivity analysis.

What do UFC fight duration trends actually show?

Rocha Santos and colleagues compared three snapshots separated by about a decade. The earliest contained 77 bouts from 10 events. The middle period had 823 bouts from 79 events. The latest had 835 bouts from 86 events.

The jump happened mainly between the first and second snapshots. Average duration increased by 140.78 seconds, or about two minutes and 21 seconds. The change from 2013-14 to 2023-24 was only 13.61 seconds.

That distinction matters. “Fights became longer” fits the 20-year comparison. “Fights are still getting longer every year” goes beyond the data.

Mean rounds tell the same story. The figure rose from 1.97 in 2003-04 to 2.36 in 2013-14. It then moved only slightly, to 2.41 in 2023-24.

The overall three-period tests were statistically significant. Pairwise tests placed the difference between 2003-04 and both modern periods. Yet the effect sizes for duration and rounds were small, at η²=0.007 and η²=0.008. Period grouping explains little of the total bout-to-bout variation.

In plain English, the era matters at group level. It won't tell you how long one matchup will last.

Are longer fights just a five-round scheduling effect?

Five-round exposure is an obvious confounder. UFC title fights can run for 25 minutes. The promotion also made non-title main events five rounds for newly signed bouts in 2011.

Modern cards therefore include more possible fight time than 2003 cards. A raw average could mistake a schedule change for a change in competition.

The researchers tested that problem. They removed five-round contests and repeated the analysis with 1,660 three-round bouts. The early period remained shorter than both modern periods.

The outcome pattern also became clearer in that restricted sample. The association between era and outcome had Cramér's V=0.151, compared with 0.061 in the full sample. That doesn't prove a cause. It does show that five-round bouts didn't create the main result.

The Unified Rules give each professional round five minutes. They cap contests at five rounds and 25 minutes. Most UFC bouts still have a 15-minute ceiling. The sensitivity test compares those like with like.

Have UFC knockouts become less common?

They were less common in the two modern snapshots than in 2003-04. KO/TKO outcomes fell from 45.5% to 32.6% in 2013-14. They stood at 34.1% in 2023-24.

Unanimous decisions moved the other way. Their share rose from 24.7% to 36.6%, then 37.1%. The full outcome distribution differed across eras, although the association was small.

This isn't a straight decline. The 2023-24 KO/TKO share was a little higher than the 2013-14 share. Those modern periods weren't statistically different from each other.

So the clean reading is a step change, followed by a plateau. The early-2000s UFC produced shorter bouts and more KO/TKO results. The two modern samples look much more alike.

An earlier longitudinal study supports the plateau idea from another angle. James and colleagues studied 2,831 men's UFC fights from 2000 through 2015. They found uneven tactical change through 2008, relative stability from 2008 to 2014, then another shift in 2015.

That study tracked performance indicators, not the same duration table. Still, both papers warn against treating every UFC era as one smooth trend line.

Why might modern UFC fights last longer?

The data establish a change, not its cause. Any explanation must stay one step behind the result.

A deeper talent pool is one plausible factor. Modern fighters usually train striking, wrestling and submissions together. Better defensive coverage can close the easy finishing routes that existed in less balanced matchups.

Tactics may matter too. Fighters who can win minutes at distance or in control positions don't need to force low-percentage finishes. Better pacing also helps them carry useful output into later rounds.

Rules and matchmaking changed across the same years. Weight classes became more settled. Regulatory oversight grew. Roster depth, event volume and the mix of divisions also shifted.

The study didn't separate those factors. It also didn't split results by weight class. That's a large gap because heavyweight fights and lighter-weight fights have different finishing patterns.

Training evidence offers context, not proof. Kirk and colleagues followed 14 MMA athletes for eight weeks. Striking drills took 20-32% of weekly training time, while MMA sparring took 2-7%. That small cohort shows structured skill work. It can't explain a UFC-wide historical change by itself.

What do Israel Adesanya, Jon Jones and Tom Aspinall show?

Individual fighters show why the era average isn't a matchup forecast.

The UFC's official record book lists Israel Adesanya at 17:09 average fight time. His UFC profile includes many scheduled five-round contests. That exposure pulls his average upward before style enters the discussion.

Jon Jones sits at 14:52 on his official profile. He also leads the UFC record book with 16 title-fight wins. Again, bout format and career level shape the number.

Tom Aspinall sits at the opposite extreme. The current UFC record book gives him the shortest average fight time among athletes with at least five UFC bouts: 2:18. His profile lists 12 knockout wins and three submissions among 15 wins.

All three competed in the modern era. Their average times still sit far apart. Division, finishing ability, opponent selection and five-round exposure can overpower the broad era signal.

Use the historical trend as a prior, not a pick. For one fight, start with scheduled rounds, division and each fighter's finish history. Then examine durability, pace and style.

What does this mean for UFC analytics?

A prediction model shouldn't treat a 2004 bout as fully exchangeable with a 2024 bout. The outcome mix changed. Training and testing across eras without a time feature can blur that shift.

The same warning applies to finish-rate baselines. A model trained on the full UFC archive may expect too many early stoppages. Recent rolling windows can fit today's environment better, though they reduce sample size.

Format needs its own variable. Three-round prelims and five-round main events don't share the same time ceiling. Title-level fighters also differ from the roster average.

Calibration matters more than a neat historical story. If a model assigns 40% to a KO/TKO, similar forecasts should land near 40% over time. The era trend can inform that estimate. It can't replace matchup data.

For bettors, longer modern averages don't create an automatic “over” angle. Sportsbook totals already reflect public information. Price, opponent interaction and uncertainty still decide whether a line offers value.

How solid is this data?

The main study used official fight footage. Three independent combat-sports analysts checked duration, ending round and method. Its 1,735 bouts cover 175 UFC events, and the three-round retest addresses the clearest format concern.

The design still has hard limits. It samples only three two-year windows, not every year from 2003 through 2024. The first window has just 77 bouts, versus more than 800 in each modern window.

Only men's fights were included. Results weren't split by division. No-contest results, disqualifications and majority draws were excluded, so the table doesn't describe every booked bout.

The study is observational. It can show that duration and result mix changed together with time. It can't isolate skill growth, rules, matchmaking or roster composition as the reason.

The small effect sizes deserve attention. Average duration changed clearly, but bouts within each era still varied widely. The standard deviations were about 358-362 seconds in all three windows.

This is useful population evidence. It remains a weak standalone tool for forecasting one fight.

FAQ

What is the average UFC fight time today?

In the study's 2023-24 men's sample, mean UFC fight duration was 650.14 seconds, or about 10 minutes 50 seconds. That average covered decided bouts across divisions and included both three-round and five-round contests.

Are UFC fights getting longer every year?

The evidence doesn't show a steady yearly rise. Fights were longer in 2013-14 and 2023-24 than in 2003-04, but the two modern periods weren't statistically different.

Are decisions more common in modern UFC fights?

Decisions were more common in the modern samples than in 2003-04. The study reported unanimous decisions at 37.1% in 2023-24, plus a separate share of split decisions.

Did five-round main events cause the increase?

Not by themselves. The researchers repeated their analysis on 1,660 fights limited to the three-round format. The early-2000s sample still had shorter fights and a different outcome mix.

Does a longer average fight time predict the over?

No. An era average describes a large group, not one matchup or a betting price. Scheduled rounds, division, styles, durability and the market line still matter.