In three independent studies, free cortisol levels after morning awakening were repeatedly measured in children, adults and elderly subjects (total n=152). Cortisol was assessed by sampling saliva at 10 or 15 minute intervals for 30-60 minutes, beginning at the time of awakening for two days (Study 1 and 2) or one (Study 3) day, respectively. In all three studies, free cortisol levels increased by 50-75% within the first 30 minutes after awakening in both sexes on all days. Premenopausal women consistently showed a stronger increase with a delayed peak after awakening compared to men on all days. In Study 2, there was a tendency for lower early morning free cortisol levels for women taking oral contraceptives (p=.10). Stability of the area under the curve (AUC) of the early morning free cortisol levels over the three (Study 1 and 2) or two (Study 3) days ranged between r=.39 and r=.67 (p<.001). Neither age, weight, nor smoking showed an effect on baseline or peak cortisol levels. Sleep duration, time of awakening and alcohol consumption also appeared to be unrelated to early morning free cortisol levels. From these data we conclude that in contrast to single assessments at fixed times, early morning cortisol levels can be a reliable biological marker for the individual's adrenocortical activity when measured repeatedly with strict reference to the time of awakening.