505 Change of Direction Test: Full Protocol and Norms by Sport

A coach hands you a stopwatch, points at a cone 5 meters away, and says run there, turn, and come back. Someone writes down 2.3 seconds. Six weeks later the same drill produces 2.28 seconds and everyone calls it progress. Neither number means much on its own, and that gap between 2.30 and 2.28 is smaller than the measurement error most stopwatch setups carry.

The 505 test itself is a good tool. It isolates a single 180-degree direction change from a short run-in, which makes it one of the cleaner field tests for tracking deceleration and re-acceleration ability over a training block. The problem is almost never the test design. It is gate placement that drifts by half a meter between sessions, warm-ups that vary from one testing day to the next, and norms pulled from a study population that has nothing in common with the athlete on the floor.

This guide sets out the protocol in enough detail that two different testers, on two different days, land within a few hundredths of a second of each other. Then it walks through what the published numbers actually say once you sort them by sport, sex, and competition level, so you can tell a real 0.1-second improvement from noise.

Why the 505 Falls Apart in Most Gyms

Stewart, Turner, and Miller (2014) tested five commonly used change-of-direction protocols, including the 505, in a controlled reliability study of 24 team-sport athletes. Under standardized conditions with electronic timing, the 505 produced an intraclass correlation coefficient (ICC) of 0.87 to 0.93 and a typical measurement error under 0.05 seconds. That is a genuinely reliable test. The catch is that number only holds when the setup is standardized. Swap in hand timing, shift the 5-meter mark by even 30cm, or let athletes choose their own pivot foot inconsistently between sessions, and the error climbs past the size of the training effect you are trying to detect.

Most training-block improvements in 505 time sit in the 0.05 to 0.10 second range over 6 to 8 weeks. If your setup carries 0.10 to 0.15 seconds of noise from inconsistent gate placement and timing method, you cannot see that improvement even when it is real. Fixing the protocol is not a formality. It is the difference between data you can act on and a number that just looks precise.

Equipment and Court Setup

Lay out the course before the athlete arrives, and measure twice. The standard 505 layout uses three marks on a straight line: the start line, a timing gate at 5 meters, and a pivot line at 10 meters.

A tape measure and a piece of chalk are enough to mark the lines, but the timing gate is not optional if you want data you can compare across sessions. A stopwatch introduces enough reaction-time variance on its own to erase a full training cycle's worth of adaptation. If dual-beam gates are unavailable, an IMU-based system that flags ground contact and re-acceleration directly from the athlete's body removes the placement-error problem altogether, since the measurement travels with the athlete rather than living in a fixed point on the floor.

Step-by-Step Testing Protocol

Run every athlete through the identical sequence below. Skipping the warm-up or shortening rest between trials is the fastest way to bias your own results toward whichever athlete happened to go first.

1. Warm-up (8-10 minutes): 3 minutes of light jogging, leg swings (10 reps per leg, sagittal and frontal plane), lateral shuffles for 2x10m, and 3 build-up sprints at roughly 60%, 80%, and 95% effort through the full 10m course. The final build-up should pass through the timing gate so the gate's trigger height is confirmed before scoring begins.
2. Practice trials: 2 submaximal attempts, one pivoting off each foot, at about 80% effort. This is where you cue technique, not where you record times.
3. Maximal trials: 2-3 full-effort attempts per leg, alternating pivot foot between trials, with 90-120 seconds of rest between each. Athletes accelerate from the start line, break the gate at 5m, contact the pivot line at 10m, turn 180 degrees, and sprint back through the gate.
4. Scoring: Record the time from the first gate break to the second gate break for every trial. Use the fastest legal trial on each leg as the athlete's score. A trial where the athlete cuts the pivot line short does not count and should be rerun after full rest.
5. Session log: Note surface, footwear, ambient temperature, and time of day. These details matter more than they seem when you retest 8 weeks later and the gym floor has been resurfaced in between.

Total testing time per athlete, including warm-up, runs about 12-15 minutes. Testing more than 6-8 athletes per hour on a single course usually means someone is cutting rest short.

How Reliable Is the 505, Really

Beyond the Stewart et al. (2014) reliability figures above, it helps to know where the error actually comes from. Nimphius, Callaghan, Bezodis, and Lockie (2018), reviewing current change-of-direction testing practice, pointed out that a large share of the variance reported across studies traces back to inconsistent operational definitions rather than to the athletes themselves: some labs use a 505 with a 15m run-in before the gate, others start cold from the 5m mark, and a few allow a rolling start. None of these are wrong exactly, but they are not the same test, and comparing raw times across them produces false conclusions about who is faster.

Practical takeaways for anyone running the test regularly:

• Typical error (the smallest change you can trust) for a well-run 505 with electronic timing sits around 0.03-0.05 seconds.
• A single session's fastest trial should not be treated as the athlete's true score if it is more than 0.05-0.08 seconds faster than the other legal trials on the same leg. That gap usually signals a timing fluke rather than a genuine outlier performance.
• Compare athletes to their own prior scores under identical setup far more than to numbers pulled from a different lab's published table.

Norms by Sport and Level

The table below pools published 505 times from studies using electronic timing and the standard 5m-plus-5m layout. Treat these as reference ranges, not pass-fail cutoffs. Surface, footwear, and exact gate placement can shift a lab's mean by 0.05-0.15 seconds even when everything else is done correctly.

Two patterns are worth acting on. First, sex differences average around 0.15-0.25 seconds across most sports, wider than the 0.05-0.10 second gaps that a good training block produces, so mixed-sex squads need separate reference bands rather than one shared target. Second, youth athletes 16-18 years old typically land 0.10-0.20 seconds slower than adult club-level peers in the same sport, which reflects incomplete strength development rather than poor technique, and should shape how aggressively you push COD-specific loading in that age group. Keep in mind that norms only tell you where an athlete sits against a population; they say nothing about why a score is low, which is where ground contact and asymmetry data become useful during the actual training block.

Errors That Quietly Wreck Your Data

Most bad 505 data does not come from a dramatic mistake. It comes from small inconsistencies that compound across a testing session.

None of these individually looks like a big deal in the moment. Stacked together across a testing day, they can easily produce more variance than eight weeks of actual training adaptation, which is exactly the scenario that leads coaches to distrust a perfectly good test.

Turning One Number Into a Training Decision

Testing without a follow-up plan just produces a spreadsheet. Once you have a clean baseline, retest on a fixed cycle, typically every 4-6 weeks during a training block, and act on the pattern rather than any single session.

• If both legs improve together and stay within 8-10% of each other, the athlete is responding to general COD training and current programming is working.
• If one leg lags and the gap widens across testing cycles, unilateral eccentric strength work on that side (single-leg Romanian deadlifts, Nordic curl progressions) usually closes the gap faster than more COD drill volume.
• If total time plateaus for two consecutive cycles despite consistent training, the limiting factor is often technique at the pivot rather than physical capacity. A short block of penultimate-step drills with video feedback, separate from strength work, tends to move the needle here.

The 505 is a simple test to run and an easy one to run badly. Standardize the setup once, log the details every time, and the number on the stopwatch starts meaning something you can actually train against.