When I first wrote this blog post in 2015, Paula Radcliffe’s 2:15:25 marathon world record stood relatively untouched at the top of the world all-time list.
Twelve years after she had set it at the 2003 London Marathon, no other woman had come within three minutes of the mark. My original post therefore consisted of a succession of increasingly breathless ways of explaining how extraordinary the performance was.
Since then, however, the statistical landscape has changed completely.
Radcliffe no longer holds the world record. Nine women have now run faster than her, while several of those athletes have produced multiple performances below 2:15:25. Following Tigst Assefa’s 2:11:04 in Berlin last month, Radcliffe now sits 10th on the world all-time list, and her 2:15:25 run is the 13th-fastest performance in history. The world record, meanwhile, has dropped to 2:09:56.
So was I wrong about Radcliffe’s run? Not really. But the more interesting question now isn’t how untouchable 2:15:25 looks. It’s how unusual it was for 2003.
A performance from a different era
Radcliffe arrived at the 2003 London Marathon already holding the world record. Six months earlier in Chicago, she had run 2:17:18, taking 89 seconds off the previous world record. Then in London she took another 1:53 off her own mark.
In the space of six months, the women’s marathon record effectively moved from 2:18:47 to 2:15:25 – an improvement of 3:22. And there wasn’t a cluster of women operating somewhere nearby.
Radcliffe won the 2003 London Marathon by more than four minutes. Catherine Ndereba, the woman who had held the world record before Radcliffe, finished second in 2:19:55.
Her splits were similarly startling. Radcliffe went through halfway in 1:08:02 and covered the second half in 1:07:23. For context: at the time, the world half marathon record stood at 1:06:44.
Radcliffe wasn’t clinging on after an overambitious first half. She was accelerating.
Put her in the men’s race
One of the comparisons from my original article still works well – provided we update the other half of it.
Radcliffe’s 2:15:25 would have placed her about 305th on the men’s world list for 2003.
Now move forward 22 years. The equivalent position on the 2025 men’s world list was roughly 2:09:14 – 42 seconds faster than the current world record.
It’s a useful illustration of what has happened to marathon running generally. A performance occupying roughly the same place in the men’s global pecking order now requires a completely different level of speed.
So simply looking at Radcliffe’s time against a modern women’s list misses an important part of the story. The event around it has changed.
And yet the women-only record still hasn’t passed it
There is another wonderfully odd statistic.
Radcliffe ran 2:15:25 in a mixed race, using male pacemakers.
World Athletics also recognises a separate world record for women competing in women-only races.
As of October 2026, that record belongs to Tigst Assefa, who ran 2:15:41 at the London Marathon in April.
In other words, 23 years later, the fastest woman ever in a women-only marathon has still not quite matched Radcliffe’s 2003 time.
There are caveats, obviously. Mixed races can provide women with benefits unavailable in women-only competition, most obviously male pacing and potentially some aerodynamic shelter. Radcliffe had that advantage in London. She started with eight male pacemakers, one of whom stayed with her for the duration of the race.
So 2:15:25 and 2:15:41 are not perfectly like-for-like performances. But the fact that the women-only record has still not overtaken Radcliffe’s old overall world record is nevertheless pretty striking.
Then the shoes arrived
The biggest reason the original version of this article now feels dated is that elite marathon running entered a completely different technological era during the second half of the 2010s.
Advanced footwear technology changed the event. Brigid Kosgei finally broke Radcliffe’s record in Chicago in 2019 with 2:14:04. Then the progression accelerated.
Tigst Assefa ran 2:11:53 in Berlin in 2023. Chepngetich ran 2:09:56 in Chicago in 2024. Fotyen Tesfay produced 2:10:51 in Barcelona in 2026. And Assefa has now lowered her own PB to 2:11:04, and had been on pace to break Chepngetich’s world record until the final two kilometres.
Training, nutrition, pacing and race organisation have evolved. More elite athletes specialise in the marathon earlier in their careers. But it would be equally silly to pretend shoes are irrelevant.
Of the 12 performances now faster than Radcliffe’s, every one was achieved many years later in a transformed marathon era. And domestically, Radcliffe is still almost six minutes clear of the next-fastest British woman, Calli Hauger-Thackery at 2:21:24.
That tells us more than repeatedly describing Radcliffe’s run as some sort of superhuman aberration.
How good was it?
When I first wrote this article, I was essentially trying to prove that 2:15:25 was one of the most astonishing performances ever produced in athletics.
A 2:15 marathon no longer exists in the statistical universe it occupied in 2003. Nine women have surpassed Radcliffe; the world record is more than five minutes faster; and performances once regarded as almost unimaginable have become possible.
But context matters. Radcliffe ran 2:15:25 in 2003. At the time, that was roughly equivalent to 305th on the annual men’s marathon list. By 2025, the same position required about 2:09:14.
Twenty-three years later, even the women-only world record remains 16 seconds slower. And after an entire technological revolution in road running, only nine women have ever run faster.
That is probably a better measure of the performance than any of the slightly giddy comparisons I came up with in 2015.
In 2003, 2:15:25 looked like a performance from the future.
Eventually, the future arrived. The interesting thing is how long it took.
Views expressed here are my own and do not necessarily represent those of any employer.