Twenty healthy athletes exercised for 30 min at four different times (beginning at 1130 and 1630) in December (darkness period) and in April (18 h of daylight). Four hours after intake of a standardized meal, a 30-min bike exercise with an intensity of 60% maximal O2 uptake was performed. Blood samples (fingertip) were drawn at 1, 5, 10, and 30 min into exercise and 5, 10, and 30 min after termination of exercise for determination of blood glucose. Glucose values were normalized by reexpressing each as a percentage of the starting value. The total area under the glucose-time curves as well as the area below the starting value was calculated. Areas were tested for the effect of sex, time of day, and season by analysis of variance. For the group as a whole during exercise, a significant effect was found by analysis of variance for sex, time of exercise, and season. During recovery, significant differences were found for sex and time of exercise but not for season. The minimal integrated glucose response to exercise occurred in females, who also showed the most rapid return to baseline values during the recovery period. Exercise in the morning produced the smallest glucose response for both sexes and faster recovery compared with exercise in the afternoon. This was also the case overall for exercise in December compared with April. This finding implies that the glycemic response may be influenced by season and timing of exercise, which may be of importance for athletes involved in vigorous training and patients with diabetes mellitus.