The Ulta Gap
Why the same races show a 1% gap and a 22% gap.
A coach reads that women have all but caught men in ultra-running. Another coach reads that men still win by more than twenty percent. They're looking at the same sport, the same year, sometimes the same races, and neither has misread the numbers. The gap in ultra-endurance isn't one figure. The argument about it is really an argument about who you count.
Get that straight and the contradictions dissolve. Nearly every widely shared claim about women closing the gap, or failing to, is measuring a different slice of the field and quietly presenting it as the whole. So before quoting a gap, it's worth knowing which runners went into it.
The median finisher and the podium are two different races
Most ultra data are hard to read because women make up only ten to thirty percent of a typical field, so the women who show up are a more filtered group than the men. Tiller and Illidi (2024) found a way around that: two ultras with roughly equal numbers of men and women, about 52 percent female. In balanced fields, the median woman and the median man finished within a percent or two of each other, and the difference wasn't statistically significant. Then they looked at the front of the same races, and the picture changed.
Figure 01 / Interactive
Who you count changes the gap
Part of the field / tap to change
In balanced fields, the median gap was 1.2% over 50 miles and 3.2% over 100. Neither reached significance.
Tiller and Illidi (2024). Through the thick middle of the field, the sexes are hard to tell apart.
Compare the top 10 of each sex in the same 50-miler and the men are 13.8% faster (p = 0.045). Over 100 miles the top-10 gap was 4.4% and not significant.
Tiller and Illidi (2024). Same race, same day, a different answer at the front.
Pool only winners and top-three athletes across 20 ultra events and the gap is 22% (95% CI 17 to 27).
Sitko et al. (2025). At the podium, women do not outperform men, and the gap is large.
One sport, three answers. Median and top-10 gaps are from two balanced-participation ultras (Tiller & Illidi 2024); the 22% is a meta-analysis of winners and top-three finishers across running, cycling, swimming and triathlon (Sitko et al. 2025). Bar length maps the gap on a 0 to 25% scale.
Why the same field gives two answers
Picture the whole field as a spread, not a single time. Most runners bunch in the middle, a few are very fast at the front, a few well off the back, and the men's and women's spreads sit almost on top of each other. Toggle between the two things a study can measure and you can watch the gap appear and vanish.
Figure 02 / Interactive
The same field, drawn two ways
Where you look / tap to change
The median sits in the overlap. Where most runners are, the two spreads nearly coincide, so the middle woman and middle man post almost the same time. Measure here and the gap looks tiny.
The podium sits in the tail. At the fast edge the men's curve reaches further right, partly from top-end physiology and partly because far more men race, giving the field more shots at an outlier. Measure here and the gap is real.
Schematic, not measured curves. Shapes are illustrative of two heavily overlapping distributions with a longer male right tail; they are drawn to explain the median-versus-podium effect, not plotted from a specific race.
So a study that reports the median tells you the sexes are nearly matched, and a study that reports the winners tells you men are well clear. Both can be true of the same afternoon, which makes a gap figure without its denominator, median or podium, close to meaningless.
The longer the race, the smaller the gap
Distance moves the number too, and in the direction the folklore predicts. In two very large race databases, Waldvogel et al. (2019) found the average sex gap shrank as races got longer, and shrank again with the athletes' age, with performance peaking around 33 for both sexes. These are averages across unbalanced fields, so they sit higher than the balanced-field medians above, but the trend inside the dataset is clean.
Figure 03 / Interactive
The gap narrows from 50 to 100 miles
Race distance / tap to change
Across 231,980 finishers over 50 miles, men were 9.13% faster on average.
Waldvogel et al. (2019). Unbalanced fields, whole-field averages.
Across 107,445 finishers over 100 miles, the average gap halved to 4.41%, and it narrowed further in older age groups.
Waldvogel et al. (2019). Peak performance sat near age 33 for both sexes.
Direction, not a fixed value. The point is the fall from 50 to 100 miles, measured the same way in the same dataset (Waldvogel et al. 2019).
Age closes it too, in fixed-distance races
The same dataset holds a second pattern that surprises most people. Step through the age groups and the average 50-mile gap shrinks from the teens onward, and in the oldest bracket the average woman actually finished ahead. That last point rests on 24 women, so treat it as a curiosity rather than a rule, but the trend beneath it is built on hundreds of thousands of finishes.
Figure 04 / Interactive
The 50-mile gap across the age groups
Age group / tap through
10–19 years
10.0%
men faster (the widest gap)
20–29 years
8.5%
men faster
30–39 years
8.1%
men faster
40–49 years
8.1%
men faster
50–59 years
8.4%
men faster
60–74 years
7.9%
men faster
75–95 years
5.6%
women faster (24 women, not significant)
Real averages, thinning samples. Mean finish times per age group across 231,980 fifty-mile records (Waldvogel et al. 2019, Table 1). The 75+ group holds only 24 women and 189 men, so its reversal wasn't statistically significant. The narrowing with age held across the fitted model.
Read the opposite way in time-limited races: Knechtle et al. (2016) found the gap in 6-hour to 10-day events grew with age rather than shrank. The measure you pick decides the answer you get.
What the physiology can and can't promise
The idea that women are built for extreme distance has a real basis. In their review, Tiller et al. (2021) set out the traits that help and the traits that cost. Which set matters depends on the event, so tap between them.
Figure 05 / Interactive
The physiology cuts both ways
What the body brings / tap to change
Works for the distance
- Greater fatigue resistanceMuscle tires more slowly under sustained effort.
- More fat oxidationLeans on fat stores, sparing limited glycogen.
- Lower energetic costOften less mass to carry over the same ground.
- Steadier pacingTends to hold a more even effort deep into a race.
These favour survival over hours, which is why the gap can shrink as races get longer.
Works against her
- Lower oxygen-carrying capacityLess haemoglobin caps the top-end aerobic ceiling.
- More gastrointestinal distressGut problems are reported more often in long events.
- Sex-hormone effectsInfluence tissue, fuel use, and injury risk across the cycle.
- Lower absolute powerLess muscle mass shows most at the fast, short end.
These bite hardest where speed and oxygen delivery decide the result, at the front of the race.
One body, two ledgers. Traits summarised from Tiller et al. (2021). The helping traits surface over extreme duration, the costing traits over speed, which is why potential and podiums can point different ways.
Physiological potential is not the same as podiums, and the places the female advantage might surface are thin on female starters. Until that changes, the elite gap tells you as much about who lines up as about who's faster.
The honest edges
Hold all of this loosely, because the evidence is thinner than the confidence around it. The balanced-participation data that make the cleanest case come from essentially two races. The meta-analysis that puts the elite gap at 22% carried very high heterogeneity between studies (above 95%), so even that figure is unstable. And women remain a minority of almost every field, roughly 26% of the 50-mile records and 20% of the 100-mile records in the Waldvogel data, close to the 20% share reported across ultra-running since 2004. That imbalance distorts every average built on it.
So the working rule for a coach, athlete or practitioner is duller than either headline and easier to defend. Before you repeat a gap number, ask which runners it counted and how many of them were women. A near-equal median doesn't promise a near-equal podium, and a 22% elite gap doesn't mean the woman next to you at the 60 km aid station is 22% behind. They're answers to different questions. The interesting work now is getting enough women to the start line to ask the question properly.
References
- Tiller NB, Illidi CR (2024). Sex differences in ultramarathon performance in races with comparable numbers of male and female finishers. Applied Physiology, Nutrition, and Metabolism 49(8):1129-1136.
- Sitko S, López-Laval I, Cirer-Sastre R, Valenzuela PL (2025). Ultra-endurance performance: are women closing the gap? A systematic review and meta-analysis. International Journal of Sports Science & Coaching.
- Tiller NB, Elliott-Sale KJ, Knechtle B, Wilson PB, Roberts JD, Millet GY (2021). Do sex differences in physiology confer a female advantage in ultra-endurance sport? Sports Medicine 51:895-915.
- Waldvogel KJ, Nikolaidis PT, Di Gangi S, Rosemann T, Knechtle B (2019). Women reduce the performance difference to men with increasing age in ultra-marathon running. International Journal of Environmental Research and Public Health 16(13):2377.
- Knechtle B, Valeri F, Nikolaidis PT, Zingg MA, Rosemann T, Rüst CA (2016). Do women reduce the gap to men in ultra-marathon running? SpringerPlus 5:672.
Link-check all references before publishing.