Football (soccer) training and matches are scheduled at different times throughout the day. Association football involves a variety of fitness components as well as psychomotor and game-related cognitive skills. The purpose of the present research, consisting of two separate studies, was to determine whether game-related skills varied with time of day in phase with global markers of both performance and the body clock. In the first study, eight diurnally active male association football players (19.1+/-1.9 yrs of age; mean+/-SD) with 10.8+/-2.1 yrs playing experience participated. Measurements were made on different days at 08:00, 12:00, 16:00, and 20:00 h in a counterbalanced manner. Time-of-day changes in intra-aural temperature (used as a marker of the body clock), grip strength, reaction times, flexibility (markers of aspects of performance), juggling and dribbling tasks, and wall-volley test (football-specific skills) were compared. Significant (repeated measures analysis of variance, ANOVA) diurnal variations were found for body temperature (p<0.0005), choice reaction time (p<0.05), self-rated alertness (p<0.0005), fatigue (p<0.05), forward (sit-and-reach) flexibility (p<0.02), and right-hand grip strength (p<0.02), but not left-hand grip strength (p=0.40) nor whole-body (stand-and-reach) flexibility (p=0.07). Alertness was highest and fatigue lowest at 20:00 h. Football-specific skills of juggling performance showed significant diurnal variation (p<0.05, peak at 16:00 h), whereas performance on the wall-volley test tended to peak at 20:00 h and dribbling showed no time-of-day effect (p=0.55). In a second study, eight diurnally active subjects (23.0+/-0.7 yrs of age) completed five test sessions, at the same times as in the first study but with a second session at 08:00 h. Test-re-test comparisons at 08:00 h for all components indicated good reliability. Intra-aural temperature showed a significant time-of-day effect (p<0.001) with mean temperature at 16:00 h (36.4 degrees C) higher than at 08:00 h (35.4 degrees C). There was no significant effect of chronotype on the temperature acrophase (peak time) (p>0.05). Diurnal variation was found for performance tests, including sit-and-reach flexibility (p<0.01) and spinal hyper-extension (p<0.05). Peaks occurred between 16:00 and 20:00 h and the daytime changes paralleled the temperature rhythm. Diurnal variation was also found for football-specific tests, including dribbling time (p<0.001, peak at 20:00 h) and chip test performance (p<0.01), being more accurate at 16:00 h (mean error=0.75 m) than at 08:00 h (mean error=1.01 m). Results indicate football players perform at an optimum between 16:00 and 20:00 h when not only football-specific skills but also measures of physical performance are at their peak. Body temperature peaked at a similar time, but positive mood states seemed to peak slightly earlier. While causal links cannot be established in these experiments, the results indicate that the diurnal variation of some aspects of football performance is affected by factor(s) other than body temperature alone.