The aim of this work was to study the effects of bright light-induced circadian phase shifts on sleep propensity and sleep architecture while the timing of the sleep/wake cycle is kept constant. Twenty-three normal subjects underwent an 11-day study including: (i) baseline sleep and vigilance evaluation; (ii) baseline evaluation of the circadian temperature rhythm with a 40-h constant routine; (iii) five hours of bright light exposure on each of three days; (iv) post-treatment sleep and vigilance evaluation; (v) post-treatment circadian rhythm evaluation with a second 40-h constant routine. Subjects were divided into three groups: eight subjects were exposed to bright light in the morning ('Morning group'), eight subjects were exposed in the evening ('Evening group'), and seven subjects were exposed in the afternoon ('Afternoon group'). After light exposure, the Morning group showed an advance of 1.23 h in the phase of the temperature rhythm, the Evening group showed a delay of 1.62 h, and the Afternoon group showed a non-significant advance of 0.5 h. In support of expectations, early-night sleep propensity was decreased by evening bright light, was increased in almost all subjects exposed to morning bright light, and was not changed by afternoon bright light exposure. The phase shift created by bright light exposure did not seem to be large enough to have a systematic effect on sleep consolidation or on REM sleep parameters in any of the three groups, suggesting that these variables are less sensitive to alterations in phase of the circadian oscillator than early-night sleep propensity.