Aims: To systematically examine the impact of prior wake, sleep dose and circadian phase on subjective alertness. Methods: Twenty-seven young males participated in one of two 12-day 28h forced desynchrony protocols varying in sleep dose (9.3h vs. 4.7h), where subjective alertness was assessed at various combinations of prior wake and circadian phase. Subjective alertness was measured using a visual analogue scale. Circadian phase was estimated using core body temperature. Results: A mixed-effects regression analysis with prior wake, circadian phase and sleep dose as fixed terms and participant as a random term revealed a sleep dose x prior wake x circadian phase interaction. Discussion: The sleep dose x prior wake x circadian phase interaction indicates that the adverse impact of sleep restriction on subjective alertness is prominent at early waking hours, particularly during the biological night.