The aims of this study were to describe and determine the test-retest reliability of an exercise protocol, the Loughborough Intermittent Shuttle Test (the LIST), which was designed to simulate the activity pattern characteristic of the game of soccer. The protocol consisted of two parts: Part A comprised a fixed period of variable-intensity shuttle running over 20 m; Part B consisted of continuous running, alternating every 20 m between 55% and 95% VO2max, until volitional fatigue. Seven trained games players (age 21.5+/-0.9 years, height 182+/-2 cm, body mass 80.1+/-3.6 kg, VO2max 59.0+/-1.9 ml x kg(-1) x min(-1); mean +/- s(x)) performed the test on two occasions (Trial 1 and Trial 2), at least 7 days apart, to determine the test-retest reliability of the sprint times and running capacity. The physiological and metabolic responses on both occasions were also monitored. The participants ingested water ad libitum during the first trial, and were then prescribed the same amount of water during the second trial. The 15 m sprint times during Trials 1 and 2 averaged 2.42+/-0.04 s and 2.43+/-0.04 s, respectively. Run time during Part B was 6.3+/-2.0 min for Trial 1 and 6.1+/-1.3 min for Trial 2. The 95% limits of agreement for sprint times and run times during Part B were -0.14 to 0.12 s and -3.19 to 2.16 min respectively. There were no differences between trials for heart rate, rating of perceived exertion, body mass change during exercise, or blood lactate and glucose concentrations during the test. Thus, we conclude that the sprint times and the Part B run times were reproducible within the limits previously stated. In addition, the activity pattern and the physiological and metabolic responses closely simulated the match demands of soccer.