Vedanta Delhi Half Marathon 2026 In Numbers

The 21st edition of the Vedanta Delhi Half Marathon was held on Sunday, 18 October 2026. Across the five timed races, 19,194 runners finished: 11,073 in the open half marathon, 7,669 in the 10K, 384 in the Police Cup and 68 elites.

As is our tradition at GeeksOnFeet, we have crunched the timing data of every finisher, and this year we have the same data for 2025 to compare against. The Great Delhi Run, the Senior Citizens Run and the Champions with Disability run are not timed and are not part of this.

How Fast Was the Field:

• The half marathon median was 2:15:20, a pace of 6:25 per km, against 2:17:45 last year. Men 2:11:47, women 2:33:12. 29.2% broke two hours, up from 26.4%, and 7.7% took over three hours, down from 10.3%.
• Three weeks earlier, at the Wipro Bengaluru Marathon on 27 September, the half marathon median was 2:39:20 and 10.6% broke two hours. At Hyderabad in August it was 2:45:27 and 7.5%. Delhi’s median runner was 24 minutes faster than Bengaluru’s and 30 minutes faster than Hyderabad’s.
• In the 10K the median was 1:20:18 against 1:23:52 in 2025, and 11.6% broke the hour, up from 7.5%. At the TCS World 10K Bengaluru 37.5% did.

	p1	p10	p25	Median	p75	p90	p99
Half Marathon 2026	1:25:20	1:45:17	1:57:57	2:15:20	2:35:51	2:55:19	3:28:15
Half Marathon 2025	1:23:20	1:46:34	1:59:05	2:17:45	2:39:38	3:00:36	3:38:16
10K 2026	45:48	58:54	1:08:16	1:20:18	1:33:10	1:46:19	2:07:46
10K 2025	45:34	1:02:28	1:12:23	1:23:52	1:38:49	1:53:30	2:17:02

The Pacing Report: Who Slowed Down:

Every half marathoner was timed at 5K, 10K, 15K and 20K, so for all 11,073 open finishers we can compare the first half with the second.

Four in five ran a positive split

• 77.9% ran the second half slower. 22.1% (2,445 runners) ran a negative split, against 22.4% in 2025.
• The median runner took 3:17 longer over the second half, about 19 seconds per km. More than one in four lost over ten percent, and 7% lost more than a fifth.
• Only 29% finished within three percent of an even split. The most even of them was Avik Pal (bib 2621): four 5K segments at 3:50 per km to the second, then 3:43 to the line for 1:20:41.

The 1:30 to 2:00 runners paced best

• 38% of sub-two-hour finishers ran a negative split. At the Tata Mumbai Marathon half in January, 16% of that group did.
• Below that the share collapses to 9.1% for the 2:30 to 3:00 runners and 4.8% for those over three hours.

How the slowdown gets worse with slower runners

The chart takes each finish group’s median pace in the opening 5K as zero and shows how many seconds per km slower each later segment was. The sub-two-hour groups barely moved and finished faster than they started. The three-hour group was running a minute and a half per km slower by 20K.

• 46.6% of the field ran their slowest 5K between 15K and 20K, almost the same as last year’s 46.3%, so this is the course. The stretch has the second pass over the Mathura Road flyover just short of 19 km.
• The finish inside the stadium pulls a sprint out of a third of the field: 34.5% ran their fastest stretch in the last 1.1 km.

Women fade less than men

• Only 13.5% of the women ran a negative split against 23.9% of the men, but the women are also slower, and slower runners fade more.
• Compared at the same finish time, the women lost less than the men in every band from 2:00 up. Among those over 3:00 the men slowed by 14.9% and the women by 11.7%.

The elites did not negative split either

• Winner Charles Matata ran 29:12 for the first half and 29:38 for the second, his slowest 5K being 15K to 20K. Mailu and Mburu, second and third, lost 33 and 46 seconds over the back half.
• Four of the 35 elite men and one of the 33 elite women ran a negative split. Wudinesh Alemu’s 33:16 and 33:16 was the exception.

The two-hour wall

• 961 runners finished in the five minutes before 2:00:00 and 733 in the five after. The 1:59 minute had 245 finishers, the busiest minute of the whole morning. It was the busiest minute last year too.
• 3,513 runners were on sub-two pace at 10K and 447 of them missed it. 17 were still on pace at 20K and lost it in the last 1.1 km. Two of them, bibs 1707 and 34048, stopped the clock at exactly 2:00:00.

The Start Line and the Overtakers:

The timing data records when every runner crossed the start line, so we can see how the field left the stadium and how many people each runner passed.

• The half marathon gun went at 5:30. 1,188 runners crossed in the first minute, then 250 to 650 a minute for half an hour. 99% were across within 31 minutes. The last starter, Rishabh Gusain, crossed an hour after the gun and still ran 2:15:09.
• That is close to the roughly 1,250 who crossed in the first minute of the TCS World 10K in April, though Delhi’s half field is two-thirds the size. It is the same first-minute crowding we flagged in Bengaluru.
• Seeding mostly worked. The fastest quarter of the field crossed at a median of 2:16 after the gun and the slowest quarter at 23:04.
• At 7:35, when the 10K went off from the same stadium, 7,971 half marathoners were still on the course. When the first 10K runner finished at 8:09, 4,180 were.

The overtakers

• The median half marathoner passed 557 runners and was passed by 516. In the 10K it was 707 and 721.
• Mir Touseef Hamid (bib 6616) started 30:27 after the gun, ran 1:40:46, and passed 6,473 half marathoners. In the 10K, Yug Agarwal (bib 45700) started 19:03 after the gun, ran 47:30 and passed 6,393.
• Those 1,188 first-minute starters had a median finish of 1:40:46, and the slowest of them took 3:50:23.

Starting at the front cost the most

• Compared at the same finish time, the runners who crossed the line in the first ten minutes faded more than those who started twenty minutes or more after the gun. Among 2:15 to 2:30 finishers, the early starters slowed by 9.4% in the second half and the late starters by 5.2%.
• The likely reason is that the late starters were held back by slower runners in the opening kilometres and passed them later, while the early starters had open road and went out too fast. 20.6% of the late 2:15 to 2:30 group ran a negative split, against 11.4% of the early group.

Who Ran: Age and Gender:

• The half marathon is capped and has settled at around 11,000, down 336 on last year. It gained 304 women and lost 640 men, so there are 4.8 men per woman against 6.1 in 2025.
• In 2025 the largest band was 35 to 44 and under-35s were 31.5% of the field. In 2026 the largest band is 25 to 34 and under-35s are 40.6%. The Gen Z shift we first saw in Bengaluru reached Delhi in a single year.
• The fastest half marathoners are in their forties. The 40 to 44 men had the fastest median of any band at 2:07:30, and the 40 to 44 women 2:27:20. The twenties are the slow decade.

The Regulars: Do Runners Learn?:

We matched 2025 and 2026 finishers by name, keeping only names that appear once in each year. 2,388 half marathoners ran both.

• Their median went from 2:13:04 to 2:11:44 and 55.8% ran faster.
• Pacing is a habit. Of the regulars who ran a negative split in 2025, 46% did it again, against 13% of those who had slowed down.

The Roll Call and Other Trivia:

• Amit edges Rahul as Delhi’s most common name. Kumar appears in 1,433 names, Singh in 1,199, Sharma in 694. There were 27 Amit Kumars.
• Together the 19,194 finishers covered 319,839 km, eight laps of the equator, and spent 36,828 hours on the course, 4.2 years of running.
• Four half marathoners finished in exactly 2:00:00, three in 2:22:22, and Francis Haokip ran the 10K in 45:00 flat.

A note on methodology:

• All times are chip times.
• The data is from mytimes.run, for the five timed races in both 2026 (19,194 finishers, 204 DNFs) and 2025 (18,909 finishers, 199 DNFs). The non-timed runs are not included.
• Halves are split at 10.55 km by interpolating between the 10K and 15K mats. A negative split means the second half was faster than the first. Only runners with all four splits are included.
• Returning runners were matched on exact full name and gender, only where the name appears once across all races in each year. First-name-plus-Kumar or Singh names were excluded. Some matches will still be two different people, which is why we report medians and shares rather than individual swings.

Hope you have liked the insights. If you find any additional insights, please do write to us via email or through any of our social media channels.

[Cover image credit - Procam Intl.]