Gymnastics Vault Run-Up Sprint Test: Measuring the Final 5-Meter Acceleration

A gymnast sticks her landing, form looks clean in the air, and the D-score panel still comes back a half-point under what the coaching staff expected. Video turns up nothing obviously wrong, so the conversation drifts to the one variable nobody actually measured: how fast she was moving in the final five meters before her feet found the board. A stopwatch split over the whole run-up can look perfectly respectable while hiding a gymnast who's already decelerating into the hurdle step instead of carrying speed onto it. That's the exact window vaulting biomechanics research keeps isolating, and it's the window most gyms have no repeatable way to test.

This guide covers a field protocol for measuring peak run-up speed in the last five meters before the vault board — the segment the research literature uses to relate approach speed to vault difficulty — along with equipment options, a session protocol, benchmark speeds by sex, level, and vault family from two published studies, and how to read a gymnast's own numbers against them.

Most gyms that track run-up speed do it with a single stopwatch split over the whole run, or a radar reading grabbed somewhere mid-approach and treated as representative. Neither resolves what a gymnast is actually doing across the distance. The FIG-legal run-up tops out at 25 meters, and the way an athlete uses that distance breaks into two jobs: roughly the first 20 meters is genuine acceleration toward a ceiling speed, and the final five meters is supposed to hold that speed steady into a controlled hurdle onto the board. A full-run average blends both jobs into one number, so a gymnast who accelerates well for 18 meters and then bleeds off speed bracing for the hurdle can post the same split as one who reaches peak speed early and carries it flat through the board — two different technical pictures behind an identical stopwatch time.

Tan et al. (2023), testing 30 Chinese national-team male gymnasts, measured this gap directly: mean speed across the full run-up came out to 6.36 m/s, while mean speed in just the last five meters reached 7.87 m/s — a difference of roughly 1.5 m/s that a single 25-meter split erases completely. That gap is why the vaulting biomechanics literature stopped reporting full-run averages and started isolating a specific window right before the board.

Schärer, Lehmann, Naundorf, Taube, and Hübner (2019) settled on a precise definition analyzing 515 vaults from 407 gymnasts at the 2016 European Championships: for handspring and Tsukahara-family vaults, run-up speed is the mean speed across the 7-to-5-meter mark in front of the vaulting table — the last full running step before the hurdle onto the board. For Yurchenko-family vaults, where the gymnast turns into a round-off well before contact, the window shifts back to 10-to-8 meters, since the round-off entry starts inside that stretch and would otherwise contaminate a straight-line reading. Use the wrong window and the number that comes out isn't measuring what it claims to.

One more detail worth flagging: the final step onto the board carries its own measurement headache, because a gymnast who over-controls the hurdle decelerates on purpose right where a coach most wants a clean reading. A protocol that only logs one average for the whole window can miss that a gymnast hit peak speed a stride early and was already slowing by the time her foot found the board — which is why the protocol below tracks peak speed and a trailing split, not just a window average.

Three setups work in a normal gym. Dual-beam photocell timing gates are cheapest and most precise for a straight-line window: place one gate at the outer edge of the measurement zone (5 m or 8 m from the board contact point, depending on vault family) and a second at the matching inner mark (7 m or 10 m), 2 meters apart, at hip height and square to the direction of travel. A Doppler radar gun or laser distance meter — Schärer et al. used a laser fixed roughly 45 meters behind the runway, aimed down the centerline at chest height — gives a continuous trace instead of one gate-to-gate average, the only way to see whether a gymnast is still building speed or already decelerating inside the window. Where neither is available, a waist-mounted GPS-plus-accelerometer unit computes instantaneous speed without a fixed sightline, though its short-segment resolution runs noisier and works best averaged across several trials.

Setup Checklist

1. Mark the actual competition run-up the gymnast uses in meet conditions — a shortened practice-gym run-up won't transfer.
2. Tape the window boundaries backward from the board: 7 m and 5 m for handspring/Tsukahara, 10 m and 8 m for Yurchenko-family vaults.
3. Square the sensor to the runway — even a 10-degree misalignment underreads true speed over a window this short.
4. Run two calibration passes at submaximal effort first.

Test on a day the gymnast isn't also grinding out landing reps or full-routine run-throughs — fatigue from unrelated vault volume flattens the exact number the test is trying to isolate.

Session Steps

1. Warm up with 4-5 progressive run-ups at 70-90% effort, including one full run into the hurdle and board contact, to lock in a normal approach before recording starts.
2. Record 6 maximal-effort attempts — full vaults or run-and-hurdle-only reps onto the board — with 2-3 minutes of full recovery between each.
3. For every trial, log peak speed inside the measurement window, the window average, and the change from the prior 15-to-10-meter split into the final window.
4. Discard any trial with a visible step-pattern break, a run that drifted off the board centerline, or a mid-run stutter step — well under 15% of attempts once a gymnast is used to the setup.
5. Average the remaining trials and note the spread across them before comparing to an external benchmark; a gymnast bouncing between 7.9 and 6.8 m/s trial to trial has a consistency problem the average alone won't show.

The table below comes from Schärer et al.'s 2019 sample of 407 European Championships gymnasts (89 female elite, 106 female junior, 89 male elite, 123 male junior), measured in the 7-to-5-meter and 10-to-8-meter windows described above. Treat this as a reference range tied to one competition cycle and one measurement method, not a universal cutoff — a gymnast's own trend across a season carries more coaching weight than where a single test session lands against these numbers.

Two patterns worth noting before using this table. Male gymnasts ran 8-9% faster than female gymnasts on handspring and Tsukahara vaults, but that gap nearly disappeared on Yurchenko vaults — likely because round-off entry technique, not raw sprint capacity, dominates that particular number. And elite gymnasts outran junior gymnasts by roughly 0.1-0.3 m/s across every vault family and sex, a gap small enough that a poorly calibrated setup or an inconsistent measurement window can swallow it whole.

Two findings from these studies change how a coach should use a speed number, and both cut against the instinct to simply chase a faster split.

Schärer et al. (2019) found run-up speed correlated with D-score at r ≤ 0.80 (R² = 0.64) for female gymnasts across vault families — a strong relationship, meaning speed accounts for a majority of the variance in the difficulty a gymnast can execute. For male gymnasts the picture split by vault type: r ≤ 0.69 for Tsukahara and r ≤ 0.65 for Yurchenko vaults, but handspring vaults showed no significant correlation between speed and difficulty at all. That matters for coaching a male vaulter on handspring-family vaults specifically: past a certain floor, chasing more speed stops moving the needle, and the ceiling comes down to takeoff technique instead. The same study found run-up speed explained 56% of the variance in flight height for female gymnasts (R² = 0.56) and 47% for male gymnasts (R² = 0.47) — real relationships that still leave close to half the picture unexplained by speed alone.

Tan et al. (2023) came at it from the sprint-capacity side: in their 30-gymnast sample, 30-meter flat sprint speed correlated strongly with 25-meter vault run-up speed on handspring and cartwheel vaults (r = 0.81, p < 0.01), and jumping ground reaction force correlated with 25-meter run-up speed (r = 0.715), last-5-meter speed (r = 0.718), and flat sprint speed (r = 0.704), all p < 0.01. Together that says a gymnast's ceiling on the runway is set substantially by sprint capacity built away from the vault, and faster final-5-meter speed tends to travel with a harder board push.

A caveat: Schärer et al.'s sample is cross-sectional — one competition, not gymnasts tracked over time — so the correlations describe who was fast and difficult at that event, not proof that adding speed to one gymnast will raise her D-score. Tan et al.'s sample is 30 male gymnasts from a single national program, which limits how far the force findings generalize to female or lower-level vaulters.

Run-up speed data earns its place in a training plan only if it gets tracked as a trend rather than a once-a-quarter test day.

• Every vault session: log 3-4 attempts with gates or a wearable sensor already in place.
• Monthly: compare the final-5m average and peak-to-average gap against the prior month — a widening gap points at a developing braking habit before it shows up as a stalled D-score.
• Before a run-up mark change: re-test the same day, since moving the mark shifts the whole acceleration curve, and a change made on feel can quietly cost speed nobody notices until the next meet.
• Competition week: taper full-vault volume but keep 2-3 recorded reps to confirm the pattern hasn't drifted from baseline.

Key References

• Schärer, C., Lehmann, T., Naundorf, F., Taube, W., & Hübner, K. (2019). The faster, the better? Relationships between run-up speed, the degree of difficulty (D-score), height and length of flight on vault in artistic gymnastics. PLOS ONE, 14(3), e0213310.
• Tan, Z., Yao, X., Ma, Y., Bi, Y., Gao, Y., Zhao, Y., & Yingjun, N. (2023). Run-up speed and jumping ground reaction force of male elite gymnasts on vault in China. Heliyon, 9(11), e21914.