Time of day variations in maximal anaerobic leg exercise were studied in 23 men mean age 23 (SD 3) years. All the subjects performed two anaerobic tests (force-velocity and multi-jump tests) and those familiar with sprinting ran an additional 50-m dash (n = 16). The maximal anaerobic powers for cycling and jumping (Pcycling and Pjump) and maximal anaerobic velocity (vpeak) were performed consecutively in the same order for all tests. The force-velocity and force-power relationships were established to determine Pcycling. The flight time (tf) and the ground contact time (tc) were recorded from five consecutive jumps on a jump-ergometer to calculate Pjump. The vpeak was measured between the 35th and the 45th m during the dash-run. The test schedules were at 0900, 1400 and 1800 hours on separate days in random order. Rectal temperatures (Tre) and body mass (mb) were measured before each test. The Tre increased significantly from 0900 to 1800 hours (P < 0.001) but mb did not vary during the day (P > 0.05). The Pcycling and Pjump were higher at 1400 and 1800 hours than at 0900 hours. The differences between the morning and the afternoon reached 3% (P < 0.05) for Pcycling and 5%-7% for Pjump (P < 0.01). The time-of-day effect was significant for tf (P < 0.05) but not for tc. During the dash-run tests, the differences almost reached significance for vpeak between 0900 and 1800 hours (P = 0.0544). No significant variations were observed between 1400 and 1800 hours for cycling, jumping and running tests. A time-of-day effect in the maximal anaerobic power of cycle and multi-jump tests existed. Such variations would have pronounced effects when expressed in competitions.