How the One-Mile Run Got Faster

Photo: Josh Kerr winning the men’s 5,000 meters at the 2025 UK Athletics Championships.  Andrewnward, CC BY-SA 4.0, via Wikimedia Commons.

A mile is a distance you can picture.  Four laps of a normal outdoor track, then a few extra steps.  For a long stretch of the last century, the public question was simple.  Could a person run that far in under four minutes?

Roger Bannister did it on a windy evening in Oxford on 6 May 1954, in 3 minutes 59.4 seconds.  That number is still the one people remember.  It is no longer the fast end of the race.  On 18 July 2026, Josh Kerr ran 3:42.66 in London.  World Athletics treated that as the men’s world record, 0.47 seconds under the mark Hicham El Guerrouj had held since 1999.  On the women’s side, the official record is still Faith Kipyegon’s 4:07.64 from Monaco on 21 July 2023.

The drop from “about four and a half minutes” to those times is not one trick.  It is better tracks, honest measuring, runners whose only job is to set the pace, and a sport that slowly let women race the distance and let professionals get paid.  The milestones below are the ones that changed the story.  They are not every line in the record book.

Line chart of men’s mile world record times falling from about 4:13 in 1886 to Josh Kerr’s reported 3:42.66 in 2026
Men’s mile marks from this article’s milestone list.  Kerr’s 3:42.66 is shown as World Athletics has reported it; ratification may still be open.  The women’s official record remains Faith Kipyegon’s 4:07.64.

When a mile was a bet

Before there was a world governing body, a mile was often a show.  In Britain in the 1700s and 1800s, the craze was called pedestrianism.  That word just meant foot racing, usually for wagers and prize money, on roads, fields, and later on small tracks.  Crowds bet.  Runners used nicknames.  Times were written down, but not always with a surveyed track and a neutral watch, so early lists disagree with each other.

By the middle of the 1800s the best professional miles were in the mid-four-minute range.  One early mark that later compilers still print is Charles Westhall’s 4:28 in London on 26 July 1855.  The man who owned the distance in the betting era was the Scot William Cummings.  On 14 May 1881, at Preston, he was credited with 4:16.2.

The amateur who came for that record was Walter George, an English runner who had already been winning miles under the new amateur rules.  Amateurs were not supposed to race professionals for money, so George turned professional to meet Cummings.  Their rematch, billed as a mile of the century, was at Lillie Bridge in London on 23 August 1886.  George won in 4:12¾, also written 4:12.8.  Cummings folded near the end.  World Athletics still points to that race as the first mile under 4:20 that the later record story has to reckon with.  It stood for nearly 30 years.

1884 Vanity Fair caricature of British miler Walter George in a dark running kit, captioned The Champion of champions
Walter George in Vanity Fair, 25 October 1884, two years before the Lillie Bridge mile.  Caricature by Carlo Pellegrini (“Ape”).  Public domain, via Wikimedia Commons.

Amateurs, a federation, and a record you could check

The amateur side grew up beside the betting side.  England’s Amateur Athletic Association held a championship from 1880, and George was its first mile winner.  Clubs measured tracks.  Newspapers printed lap times.  A “world record” in that era still meant “the best time serious statisticians would accept,” not a certificate from a world office.

That office arrived in 1912, as the International Amateur Athletic Federation.  It later shortened its name to the IAAF, and it is now World Athletics.  In 1913 it began recognizing an official mile record.  The first one was John Paul Jones of the United States, 4:14.4, at Allston, Massachusetts, on 31 May 1913.

Two years later Norman Taber ran 4:12.6, also in the Boston area, on 16 July 1915.  He had help from pacemakers who started ahead of him on handicaps, which is a messy way to set a record, and the mark only nudged George’s old professional time.  The cleaner breakthrough was Paavo Nurmi of Finland, 4:10.4 in Stockholm on 23 August 1923.  Nurmi was the great timer of his age.  He raced with a stopwatch in his hand and made the mile look like a plan instead of a scramble.

Through the 1930s the record crept down in small bites, on cinder tracks.  Cinders are the crushed ash and clay surface almost every big track used then.  Jules Ladoumègue, Jack Lovelock, Glenn Cunningham, and Sydney Wooderson each took a piece off it.  By 1937 the record was Wooderson’s 4:06.4.  Newsreels made these men famous.  The mile was already a public clock, not only a club race.

Nine years stuck just past four minutes

Then the war narrowed who could race.  In neutral Sweden, Gunder Hägg and Arne Andersson spent the early 1940s taking the record off each other.  Hägg’s third version, 4:01.4 at Malmö on 17 July 1945, was the one that stuck.  Under the rounding rules of the time, a hand time — a watch stopped by a person, not by an electronic beam — was moved up to the next fifth of a second.  Hägg’s watches had caught something a tick faster.  The ratified record was 4:01.4.

That mark lasted nine years.  Runners got close.  None of them got under four minutes in a race the federation would accept.  The barrier became a story people told, half sporting and half superstitious, as if four minutes were a wall in the body rather than a round number on a clock.

Bannister, then Landy, then a race with two men under four

Roger Bannister was a 25-year-old medical student.  The attempt at Oxford’s Iffley Road track, on 6 May 1954, was almost called off.  A crosswind was blowing, with gusts strong enough to make a record look foolish.  He ran anyway, in an Amateur Athletic Association match against the university, in front of about 3,000 people.

Two friends did the pacing.  A pacemaker is a runner whose job is to carry the speed, then step aside.  Chris Brasher led the first part.  Chris Chataway took over.  Bannister went past on the last bend.  The first lap was about 57.5 seconds.  The last lap was about 58.9.  The time was 3:59.4.  Norris McWhirter’s announcement drowned in the noise before he finished the tenths.  Bannister already knew.  He said later that he felt like an exploded flashlight, and that he knew before he heard the time.

Blue plaque on a brick wall beside Iffley Road in Oxford, marking Roger Bannister’s sub-four-minute mile on 6 May 1954
The plaque is on the wall by Iffley Road, not a photograph of the 1954 race.  The track that day was cinders.  Photo by Owen Massey McKnight, CC BY-SA 2.0, via Wikimedia Commons.

The wall did not stand for a sporting lifetime.  It stood for 46 days.  On 21 June 1954, in Turku, Finland, Australia’s John Landy ran 3:57.9 on the watches.  The rule then rounded a record up to the next fifth of a second, so the ratified mark was 3:58.0.  When the federation later switched to tenths, it did not go back and rewrite Landy’s line.

They met on 7 August 1954 in the mile at the British Empire and Commonwealth Games in Vancouver.  It was the first race in which two men broke four minutes.  Landy led.  Near the end he looked over his left shoulder.  Bannister passed on the right and won in 3:58.8, with Landy second in 3:59.6.  People still call it the Miracle Mile.  Bannister had the first sub-four.  Landy had the record.  Neither fact cancelled the other.

From a barrier to a moving target

Once four minutes was ordinary for the best men, the record became a chase again.  Derek Ibbotson ran 3:57.2 in London in 1957.  Herb Elliott of Australia took a real chunk out of it on 6 August 1958 in Dublin, 3:54.5.

Peter Snell of New Zealand matched that kind of calm in an unlikely place.  On 27 January 1962, on a grass track at Wanganui, he ran 3:54.4.  He improved it to 3:54.1 in Auckland in 1964.  Michel Jazy went to 3:53.6 in 1965.  Then Jim Ryun, an American who was still 19 when he first broke the record, ran 3:51.3 at Berkeley in 1966 and 3:51.1 at Bakersfield on 23 June 1967.

The next clear shift in style came from Filbert Bayi of Tanzania.  On 17 May 1975 in Kingston, Jamaica, he led from the gun and ran 3:51.0.  Front-running a mile record is rare, because the leader eats the wind.  Less than three months later, on 12 August 1975 in Gothenburg, John Walker of New Zealand ran 3:49.4.  That was the first mile under 3:50.

A few of the men’s marks that carry the story, official times as ratified:

  • 4:12¾, Walter George, London, 23 August 1886.  The professional mile that later eras had to beat.
  • 4:14.4, John Paul Jones, Allston, 31 May 1913.  First mile record the international federation recognized.
  • 4:10.4, Paavo Nurmi, Stockholm, 23 August 1923.
  • 4:01.4, Gunder Hägg, Malmö, 17 July 1945.  The mark that sat for nine years.
  • 3:59.4, Roger Bannister, Oxford, 6 May 1954.  First ratified mile under four minutes.
  • 3:58.0, John Landy, Turku, 21 June 1954.  Watches had 3:57.9.  Rounding made the official line 3:58.0.
  • 3:54.5, Herb Elliott, Dublin, 6 August 1958.
  • 3:51.1, Jim Ryun, Bakersfield, 23 June 1967.
  • 3:49.4, John Walker, Gothenburg, 12 August 1975.  First under 3:50.
  • 3:46.32, Steve Cram, Oslo, 27 July 1985.
  • 3:43.13, Hicham El Guerrouj, Rome, 7 July 1999.  Noah Ngeny was second in 3:43.40 in the same race.
  • 3:42.66, Josh Kerr, London, 18 July 2026.  The mark World Athletics has reported as the current record.

Coe, Ovett, and a record that changed hands in a weekend

By the late 1970s the mile’s theater was British again, and it was personal.  Sebastian Coe ran 3:48.95 in Oslo on 17 July 1979, ratified as 3:49.0 because hand timing was still the official standard for that record.  A year later, on 1 July 1980, also in Oslo, Steve Ovett took it with 3:48.8.  Ovett won the Olympic 800 that summer.  Coe won the Olympic 1,500.  The mile was a separate argument.

In August 1981 they traded it twice in nine days.  Coe ran 3:48.53 in Zürich on 19 August.  Ovett ran 3:48.40 in Koblenz on 26 August, one of the shortest reigns in the event’s history.  Two days later, on 28 August in Brussels, Coe ran 3:47.33.  Fully automatic timing, to the hundredth of a second, was the official standard for these longer track races from 1981.  The record stopped being a watch in someone’s hand.

Steve Cram, younger than both, took the next big slice at the Bislett Games in Oslo on 27 July 1985, in 3:46.32.  Oslo’s mile was already called the Dream Mile, because records kept falling there.  The surface was no longer cinder.  Synthetic tracks, rubber instead of ash, had been the Olympic standard since Mexico City in 1968.

El Guerrouj, and 27 years of 3:43

Noureddine Morceli of Algeria broke Cram’s record by almost two seconds.  On 5 September 1993 in Rieti he ran 3:44.39.  That looked, for a while, like a number that would sit.  Hicham El Guerrouj of Morocco did not leave it there.  On 7 July 1999, at the Golden Gala in Rome’s Stadio Olimpico, he ran 3:43.13.  The man on his shoulder, Kenya’s Noah Ngeny, ran 3:43.40, which was itself one of the fastest miles ever run.  The track in Rome was known as a quick one.  The time was still his.

El Guerrouj’s mile lasted 27 years.  He was not only a miler.  On 14 July 1998, also in Rome, he had run 3:26.00 for 1,500 meters, and that mark was still the men’s 1,500 record on World Athletics’ lists in 2026.  At the Athens Olympics in 2004 he finally won the 1,500, then the 5,000, after earlier global finals that had gone wrong.  The Olympic race is the shorter one.  The mile is the one the English-speaking public kept.

Kerr, London, and a mile faster than Bannister’s 1,500

Josh Kerr’s attempt had a nickname before it had a time.  Project 222 meant 222 seconds, or 3:42.00, a round target just under El Guerrouj’s 3:43.13.  He did not need all of that.  On 18 July 2026, in the Emsley Carr Mile at the London Diamond League meet, the pro track-meet series, he ran 3:42.66.

The pace was shared, then it was his.  Brannon Kidder and Žan Rudolf were the designated pacemakers.  Kerr went through 400 meters in 55.3, 800 in 1:51.1, and 1,200 in about 2:46.5.  His 1,500-meter split was 3:27.62, faster than his own British record for a full 1,500, and the last lap was still the race.  Yared Nuguse finished second in 3:45.69.  Fast, and still behind El Guerrouj’s 3:43.13 and Ngeny’s 3:43.40 on the all-time list.  Kerr said afterward that he was deaf in the last 110 meters.

World Athletics’ own write-up of the race makes a comparison that is hard to improve on.  Kerr’s full mile, 3:42.66, was faster than the 3:43.0 Bannister was timed at for 1,500 meters during the 1954 race.  The whole mile, in 2026, fit inside what used to be the split on the way to four minutes.

A small caution belongs on the line, because new marks sometimes do.  Diamond League race reports often print an asterisk while a record is still going through formal ratification.  World Athletics’ feature on the race, and its 2026 season list, already present 3:42.66 as the world record.  This article follows that reporting.  It is not a substitute for the ratification notice if you are checking the record book later.

The women’s mile was a best time before it was a record

Twenty-three days after Bannister, on 29 May 1954, Diane Leather ran 4:59.6 in Birmingham, at the Alexander Sports Ground.  She was the first woman under five minutes.  She had no pacemakers.  She had already run an 800 meters earlier that afternoon.  Three days before, on 26 May, she had stopped at 5:00.2.  Reports from the day have her sounding relieved.  She still had chemistry exams.

British Pathé’s reel is that near-miss, the run just over five minutes, not the 4:59.6 itself.  The footage is still the closest public look at the attempt.  Leather then took the best time down to 4:45.0 in London on 21 September 1955.  It lasted as a world best for more than seven years.

It was a world best, not a world record, for a blunt reason.  The international federation did not keep an official women’s mile record until 1967.  Women’s distance racing had been held back in the rulebook for decades.  After the 1928 Olympics, officials dropped the women’s 800 meters and did not put it back until 1960.  The women’s Olympic 1,500 meters did not exist until Munich in 1972.  Leather could break five minutes and still be outside the events the sport treated as serious.

Anne Smith of Britain supplied the first official women’s mile record, 4:37.0 in London on 3 June 1967.  Mary Decker, later Mary Slaney, ran 4:21.68 in Auckland in 1980 and 4:16.71 in Zürich on 21 August 1985.  Paula Ivan ran 4:15.61 in Nice in 1989.  Svetlana Masterkova ran 4:12.56 in Zürich on 14 August 1996, and that number sat into the next century.

Kipyegon, and a faster mile that does not count

Sifan Hassan moved the official record by a small, hard amount.  On 12 July 2019, in Monaco, she ran 4:12.33.  Faith Kipyegon did not chip at it.  On 21 July 2023, on the same track, Stade Louis II, she ran 4:07.64.  World Athletics ratified it.  She won by almost seven seconds.  It was her third world record in about seven weeks, after a 1,500 under 3:50 and a 5,000.  The mile was the number people outside the sport already understood.

Faith Kipyegon kissing a medal beside a red track at the 2022 World Athletics Championships
Kipyegon after the 1,500 meters at the 2022 World Championships in Eugene.  That is the Olympic distance, about 109 meters short of a mile.  Photo by Erik van Leeuwen, GFDL, via Wikimedia Commons.

On 26 June 2025, at Stade Charléty in Paris, she ran faster.  The clock read 4:06.42.  She had been aiming at four minutes.  She went through 400 meters in 1:00.20, 800 in 2:00.75, and 1,200 in 3:01.84, then closed in about 63 seconds.  World Athletics was explicit.  This was an exhibition, set up as Nike’s Breaking4 project.  It was not eligible for a world record.  Men were in the pacing group.  The kit and shoes were part of a specially built attempt.  The time is real as a performance.  It is not the record.

The official record, checked against World Athletics’ all-time list on 2 October 2026, is still 4:07.64.  The next official mile is Gudaf Tsegay’s 4:11.88, set in London on 19 July 2025.  Hassan’s 4:12.33 is third.  A women’s mile at the 2026 Prefontaine Classic did not replace the record.  The fast end of that race, on the all-time list, sits back in the 4:17s.

Selected women’s marks, with the same caution.  Early lines are best times.  From 1967 they are official records, except the Paris exhibition.

  • 4:59.6, Diane Leather, Birmingham, 29 May 1954.  First woman under five minutes.  Not an official world record.  No pacemakers.
  • 4:45.0, Diane Leather, London, 21 September 1955.  Stood as a world best for more than seven years.
  • 4:37.0, Anne Smith, London, 3 June 1967.  First women’s mile the federation recognized.
  • 4:16.71, Mary Slaney, Zürich, 21 August 1985.
  • 4:12.56, Svetlana Masterkova, Zürich, 14 August 1996.
  • 4:12.33, Sifan Hassan, Monaco, 12 July 2019.
  • 4:07.64, Faith Kipyegon, Monaco, 21 July 2023.  The official record as of this writing.
  • 4:06.42, Faith Kipyegon, Paris, 26 June 2025.  Faster, and not record-eligible.  The official book did not change.

Why the Olympics almost never run a mile

A statute mile is 1,609.344 meters.  The race you see at the Olympics and at most world championships is 1,500 meters.  The gap is 109.344 meters, a bit more than the length of a 100-meter straight, plus a few strides.  Four laps of a 400-meter track are only 1,600 meters, so a full mile is about nine meters longer than four laps.  On a track, the mile start is set back from the usual 1,500 start so the finish line can stay put.

At the speed of these records, that extra 109 meters is worth roughly 15 to 17 seconds.  It is not a magic conversion.  A tiring runner loses more than a fresh one.  El Guerrouj’s own pair is a clean illustration, not a formula.  His 1,500 record is 3:26.00.  His mile was 3:43.13.  The difference is a little over 17 seconds.  Kipyegon’s ratified 1,500 record is 3:48.68, set at the Prefontaine Classic in Eugene on 5 July 2025.  Her official mile is 4:07.64.  Same rough gap.  Since 1976, the mile has been the only non-metric distance World Athletics still keeps for record purposes.  Everything else on the books went metric.  The mile stayed because people already cared about it.

Faith Kipyegon celebrating on a blue track with a Kenyan flag after the 2016 Olympic 1500 meters
Kipyegon after winning the Olympic 1,500 meters in Rio in 2016.  The Games schedule the metric race, not the mile.  Photo by Erik van Leeuwen, GFDL, via Wikimedia Commons.

What actually moved the clock

George and Bannister ran on cinders.  From the late 1960s the records moved onto synthetic rubber, and a hand watch with a rounding rule gave way to a fully automatic clock.

Pacing changed the shape of the race.  Bannister had two friends doing it in the open.  Leather did not.  A modern Diamond League mile hires pacemakers, and some meets add a rail of lights, often called Wavelight, that shows record pace.  The lights do not finish the race.

The other change is who was allowed to race, and to be paid.  Amateur rules kept George from meeting Cummings without turning professional, and they kept the women’s mile off the official books.  From the 1980s athletes could be paid in the open.  Newer racing shoes, with a stiff plate and thick foam, are part of this era’s equipment.  They are not a reason to skip the name on the result sheet.

Indoor miles, and miles on the road, are separate record lists.  A fast mile in an arena, or a road mile with a net drop, does not replace the outdoor track record this story is about.  If you only remember two outdoor numbers from the autumn of 2026, remember these.  Men, Josh Kerr, 3:42.66, London, 18 July 2026, as World Athletics has reported it.  Women, Faith Kipyegon, 4:07.64, Monaco, 21 July 2023, still the official record, with a faster Paris exhibition sitting beside it and not in the book.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top
Aglena