Anaerobic Shuttle Capacity Test Protocol: Setup, Scoring, and Team-Sport Norms

Every preseason has the same moment. The squad lines up for the 20m multistage shuttle run, the beeps get faster, and by level 13 or 14 half the roster has dropped out while a handful of engine players keep grinding to level 17. The numbers go in the spreadsheet and the staff moves on believing they now know who is conditioned. Then week three of the season arrives, and the same engine players who topped the beep test are losing a step on the fourth or fifth transition sprint of a tight second half, while a player who quit the beep test early is still winning foot races late in the game.

That mismatch is not a fluke. A continuous, incrementally-paced shuttle run measures aerobic capacity — how long the cardiovascular system can keep supplying energy at a rising, sustained pace. It says almost nothing about how well an athlete repeats short, maximal efforts separated by brief, incomplete recovery, which is the actual demand of most field and court sports. A winger does not jog at a steady pace for four minutes; she sprints 15-20m flat out, gets 15-25 seconds to half-recover, and does it again, dozens of times a match. That is a different energy system and a different fatigue mechanism, and it needs its own test. The protocol below builds a short-distance, incomplete-recovery anaerobic shuttle test, walks through setup, scoring math, and norms from the repeated-sprint literature so you can tell a genuine anaerobic capacity deficit apart from a slow day.

What Your Beep Test Score Is Hiding

Bishop, Girard, and Mendez-Villanueva (2011), in the second half of their two-part Sports Medicine review of repeated-sprint ability, made a point most performance staff eventually learn the hard way: across the studies they reviewed, aerobic fitness markers like maximal oxygen uptake showed only a weak-to-moderate relationship with how much an athlete's sprint times fall off across repeated efforts, with correlations generally sitting in the 0.3-0.5 range rather than anywhere close to 1.0. Knowing an athlete's beep test level tells you comparatively little about whether they can hold their speed through rep six of a repeated-sprint sequence. The review's own stated limitation matters too: the field has never converged on a single standardized test, so shuttle distance, rep count, and recovery duration vary widely study to study, which makes it risky to treat any published norm as a universal cutoff rather than a reference point for your own protocol.

The takeaway is not that aerobic conditioning is worthless. It is that a continuous, progressively-paced shuttle test and a short, maximal, incomplete-recovery shuttle test measure genuinely different qualities, and a roster can be strong on one while weak on the other. Testing only the aerobic side leaves a real, coachable deficit invisible until it surfaces as a late-match sprint that never quite arrives.

Equipment and Lane Setup

The test needs a short, flat shuttle lane with a clearly marked turn line and a way to capture split times for six reps without the athlete reading a stopwatch mid-effort.

Surface and footwear matter more here than in a straight-line sprint test, since every rep includes a full deceleration, plant, and re-acceleration at the turn line. Switching from rubber turf to a hard court between sessions can shift turn times by several tenths of a second independent of any real change in anaerobic capacity, so lock the surface and log it every time.

Step-by-Step Testing Protocol

1. Warm-up (10-12 minutes): Light jog, dynamic mobility, then two build-up shuttle reps at roughly 70% and 90% effort, each finished with a full stop at the turn line.
2. Set the shuttle distance: Mark a 10m lane. Each rep is 20m total: 10m out, a touch-and-turn at the far line, 10m back to the start/finish line.
3. Familiarization: One submaximal rep at about 85% effort, focused on a clean, controlled turn rather than a stumbled pivot.
4. Maximal reps: The athlete runs 6 reps of the 20m shuttle at full effort, starting each rep from a static, athlete-chosen stance behind the start line.
5. Recovery between reps: 20 seconds of passive or slow-walk recovery, timed from the finish of one rep to the start of the next. Call the start of each rep on the clock rather than letting athletes self-pace their recovery.
6. Capture the numbers: Record each rep's time to the nearest 0.01-0.1 second. A rep only counts if the athlete's foot fully crosses or touches the turn line before reversing direction.
7. Between-athlete rest: For group testing, allow at least 8-10 minutes of full recovery before an athlete repeats the sequence on a retest day.

The six-rep block takes about 3 minutes of work-and-rest given roughly 3.3-4.2 seconds per rep and 20-second recovery windows. That short footprint is deliberate: a work:rest ratio near 1:5 to 1:6 keeps each rep close to maximal and prevents aerobic contribution from masking the anaerobic signal, following the short-effort, incomplete-recovery logic Fitzsimons, Dawson, Ward, and Wilkinson (1993) used when they built one of the first shuttle-format repeated-sprint tests to separate anaerobic capacity from the continuous-running tests already common at the time.

Scoring: Best Time, Total Time, and Fatigue Index

Three numbers come out of the six reps, and they answer different questions.

Best Time (BT) is simply the fastest single rep, usually rep one or two. It reflects raw shuttle speed more than anaerobic capacity on its own.

Total Sprint Time (TST) is the sum of all six rep times. It reflects overall output across the set.

Fatigue Index (FI), sometimes called percent decrement, measures how much performance drops off relative to what a perfectly repeated best effort would have produced: FI (%) = ((TST − (BT × 6)) / (BT × 6)) × 100.

Worked example: an athlete records rep times of 3.42s, 3.48s, 3.55s, 3.60s, 3.65s, and 3.72s. Best Time is 3.42s. Total Sprint Time is 21.42s. The ideal total, if every rep matched the best time, would be 3.42 × 6 = 20.52s. Fatigue Index works out to ((21.42 − 20.52) / 20.52) × 100 = 4.4%. A second athlete with the same 3.42s best time but rep times drifting out to 3.42, 3.55, 3.70, 3.85, 3.95, and 4.05 posts a Total Sprint Time of 22.52s against the same 20.52s ideal, a Fatigue Index of 9.7% — more than double, despite an identical opening rep. Best Time alone would have called these two athletes equal.

What the Research Actually Shows

Spencer, Fitzsimons, Dawson, Bishop, and Goodman (2006), publishing in the Journal of Science and Medicine in Sport, ran a repeated shuttle-sprint protocol on field hockey players and found Best Time highly reliable session to session, with a coefficient of variation around 1.5% — small enough to trust a single session's reading. Fatigue Index told a different story: its coefficient of variation ran well into the 13-20% range, meaning a single session's percent-decrement score can swing widely on noise alone, independent of any real change in condition. The authors were direct about the limitation this creates: Fatigue Index should not be treated as a precise, standalone score from one session, and is far more useful averaged across sessions or tracked as a longer-term trend.

Fitzsimons, Dawson, Ward, and Wilkinson (1993), publishing in the Australian Journal of Science and Medicine in Sport, were among the first to formalize this style of test, building a running-based repeated shuttle-sprint protocol using short, near-maximal efforts and brief incomplete recovery specifically to isolate anaerobic capacity from the continuous-running and continuous-shuttle protocols that dominated fitness testing at the time. Their design logic — short work bouts, incomplete recovery, six-plus repetitions — is the direct basis for the protocol above. The limitation they flagged remains relevant three decades later: performance across reps is influenced by pacing strategy as well as true anaerobic capacity, so an athlete who goes out conservatively on rep one produces a flatter, better-looking Fatigue Index without necessarily having superior conditioning — which is why enforcing an all-out effort from rep one matters as much as the distances and timing.

Norms and How to Read Them

The bands below are drawn from the range of Fatigue Index values reported across repeated-sprint studies using short shuttle or linear sprints with incomplete recovery. Given the reliability limitation above, treat these as a starting classification, not a pass-fail line, and weigh them behind the athlete's own multi-session trend.

Two comparisons matter more than the band an athlete lands in. Compare Best Time separately from Fatigue Index: fast Best Time with a high Fatigue Index means raw speed without repeat-effort capacity — a specific, trainable gap, not a general speed problem. And track Fatigue Index across sessions rather than anchoring to one test day; the Spencer et al. (2006) data suggests a 2-3 percentage point swing on a single retest is normal noise, while a sustained shift across three or four sessions is the signal worth acting on.

Mistakes That Wreck the Score

What to Do With a Poor Fatigue Index

A Fatigue Index above 8-12% is a training target, not a conditioning verdict on the whole athlete. Athletes in that range typically respond to a block built around short, maximal repeated efforts at work:rest ratios close to what the test itself uses — 15-20m sprints with 15-20 seconds of recovery, building from 4-5 reps toward 8-10 across a training block — rather than more continuous aerobic running, which trains a different system almost entirely. Phosphocreatine resynthesis between efforts is the limiting factor for most team-sport athletes in this range, and that responds to repeated short-effort exposure far faster than it responds to added distance running.

If Best Time is strong but Fatigue Index is poor, train the drop-off directly rather than chasing raw speed work the athlete likely does not need. If both are weak, fix sprint mechanics first before loading repeated-effort volume, since poor technique under fatigue tends to compound rather than average out. Retest every 4-6 weeks rather than every week; the Spencer et al. (2006) reliability data means week-to-week Fatigue Index swings are largely noise, and chasing them means adjusting a program based on measurement error.