Total sleep deprivation (TSD) in humans is associated with altered hormonal levels, which may have clinical relevance. Less is known about the effect of an extended sleep period before TSD on these hormonal changes. Fourteen subjects participated in two experimental counterbalanced conditions (randomised cross-over design): extended sleep (21.00–07.00 h time in bed, EXT) and habitual sleep (22.30–07.00 h time in bed, HAB). For each condition, subjects performed two consecutive phases: six nights of either EXT or HAB. These nights were followed by 3 days in the sleep laboratory with blood sampling at 07.00 and 17.00 h at baseline (B-07.00 and B-17.00), after 24 and 34 h of continuous awakening (24 h-CA, 34 h-CA) and after one night of recovery sleep (R-07.00 and R-17.00) to assess testosterone, cortisol, prolactin and catecholamines concentrations. At 24 h of awakening, testosterone, cortisol and prolactin concentrations were significantly lower compared to B-07.00 and recovered basal levels after recovery sleep at R-07.00 (P < 0.001 for all). However, no change was observed at 34 h of awakening compared to B-17.00. No effect of sleep extension was observed on testosterone, cortisol and catecholamines concentrations at 24 and 34 h of awakening. However, prolactin concentration was significantly lower in EXT at B-07.00 and R-07.00 compared to HAB (P < 0.05, P < 0.001, respectively). In conclusion, 24 h of awakening inhibited gonadal and adrenal responses in healthy young subjects and this was not observed at 34 h of awakening. Six nights of sleep extension is not sufficient to limit decreased concentrations of testosterone and cortisol at 24 h of awakening but may have an impact on prolactin concentration.