HYROX Sled Pull: weights, rules and time benchmarks

Sled Pull needs its own benchmark: your division, comparison group and event conditions matter. The sled's load alone does not explain the whole performance.

This guide uses 4,985 participations across five Brazilian events and 212,357 across 70 international events, after the filters explained below. We reopened 23 extreme-split records at the official source and excluded them from the benchmarks. The figures describe the available archive; they are not guaranteed targets or certification of each athlete's execution.

Sled Pull weights and rules

It is the third station, after the third run. Singles complete 50 m pulling the sled, in four 12.5 m lengths. Total loads include the sled:

Remain standing in the designated area and keep the rope in your lane. The entire sled must cross the line before reversing direction. Check standards and penalties in the official 2026/27 rulebook, section 8.3, reviewed on 25 October 2026. These loads do not automatically apply to doubles or other formats.

How to interpret the benchmarks

Times use minutes:seconds. The fastest 25% benchmark separates roughly the fastest quarter of participations; the group midpoint divides the sample into two halves. The first benchmark and upper boundary enclose approximately the central half.

A faster result than the first benchmark stands out in that station and dataset. A time above the central range invites investigation; it does not establish insufficient strength, a technical error or recoverable seconds. Ties and rounding slightly affect proportions.

Brazil and international: initial benchmarks

These are race locations, not nationalities. Brazil covers 20 September 2025–18 October 2026; international data cover 19 July 2025–9 August 2026. Windows and composition differ. The international extraction includes only the 70 completed event extracts; 29 planned events are absent. This is not the entire database or a representative worldwide sample.

Brazil

International

Women Pro requires particular care: Brazil's midpoint is 5:50; removing historically provisional records changes it to 6:12. Internationally it changes from 5:55 to 6:22. Flags may be outdated, and a missing flag does not prove official confirmation. This sensitivity shows how sample composition affects a benchmark. The organiser may revise source publications.

A Brazil–international difference does not establish athlete superiority or measure surface friction. Prefer your own event for individual review where coverage permits.

Age-group benchmarks

Read each time cell as group midpoint / fastest 25% benchmark. Participation and event counts appear alongside. Two benchmarks require at least 50 participations across three events. With 20 across two events, only the midpoint appears, with an asterisk. Below that, no benchmark is shown. These are editorial support criteria, not confidence percentages.

Women Open by age

Men Open by age

Women Pro by age

Men Pro by age

Six Brazilian and 50 international participations had no interpretable age and enter only the overview. Age bands follow the stored result. Missing support does not mean a category does not exist.

A real example: 5min16s in Sled Pull

In the previously verified São Paulo October 2026 case, a Men Open participation aged 35–39 recorded 5:16. Among 245 participations from the same event, sex, division and age with complete race splits, the midpoint was 4:51, an observed gap of 25 seconds.

That subgroup requires the whole race to be complete; the tables instead require a Pull split and eligible finish. The official detail confirms the example, not every record in this new dataset. See Beyond the finish time.

Use the gap to ask about pauses, rope handling, penalties or remembered conditions. The split cannot identify the answer or promise those 25 seconds can be recovered. For pushing, use the separate Sled Push benchmarks.

Sources, checks and limitations

Primary sources: official results for 2025/26 and 2026/27, reviewed on 25 October 2026. Checking the 23 detail records is not an individual revalidation of the entire archive.

• Frozen dated-event scope from seasons 2025/26 and 2026/27, restricted to Men/Women Open/Pro singles. Previously reconciled event identity and sex, unresolved-edition exclusions and removal of Overall/Week/Elite aggregate views were retained. These are participations, not unique people.
• Eligible finishes require source identification, rank, no blocking integrity flag and a time from 45 minutes to five hours. Pull requires exactly one positive station split, order 6 in the alternating run/station archive, below finish time and no station penalty marked in available fields. Missing flags do not prove no penalty occurred.
• Of 4,987 eligible Brazilian finishes, one lacked a valid split. Of 212,404 international finishes, 22 had invalid splits and three had splits at or above finish time. That initially left 4,986 and 212,379 participations. An additional 30-second to 30-minute station screen removed one Brazilian and 22 international records, producing the published totals.
• Quartiles were calculated from second-level frequencies, without averaging event medians. Four general Brazilian distributions were also checked with an independent local calculation. Every official record was not revalidated.
• Extreme-value review: we opened the 23 detail records with splits below 30 seconds or above 30 minutes. Twenty-two stored station values appear in the source; one Sled Pull split is unavailable in its current detail record. Six records display explicit Sled Pull penalties missing from the extracted flags. All 23 were excluded from benchmarks, without changing the database, imputing missing times or subtracting penalties. The interval is an analytical quality screen, not a declaration that excluded participations are invalid. Restoring all 23 changes no general midpoint and changes displayed age-group midpoints by at most one second.
• Surface, friction, equipment, technique, experience, prior effort and individual conditions are unmeasured. The analysis is descriptive, with no physiological diagnosis or Open/Pro conversion.

Find your participation

Open the results hub and choose an available race. Search and the available preview are free, without registration at this stage. The full assessment is paid, with contents, price and availability shown before purchase. Official rankings remain the organiser's responsibility; OUTLIER provides independent interpretation.