Sleep deprivation induced by cuff inflations during overnight blood pressure (BP) monitoring might interfere with the prognostic significance of nighttime BP. In 2934 initially untreated hypertensive subjects, we assessed the perceived quantity of sleep during overnight BP monitoring. Overall, 58.7%, 27.7%, 9.7%, and 4.0% of subjects reported a sleep duration perceived as usual (group A), <2 hours less than usual (group B), 2 to 4 hours less than usual (group C), and >4 hours less than usual (group D). Daytime BP did not differ across the groups (all Ps not significant). Nighttime BP increased from group A to D (124/75, 126/76, 128/77, and 129/79 mm Hg, respectively; all Ps for trend <0.01). Over a median follow-up period of 7 years there were 356 major cardiovascular events and 176 all-cause deaths. Incidence of total cardiovascular events and deaths was higher in the subjects with a night/day ratio in systolic BP >10% compared with those with a greater day-night BP drop in the group with perceived sleep duration as usual or <2 hours less than usual (both P<0.01), not in the group with duration of sleep >or=2 hours less than usual (all Ps not significant). In a Cox model, the independent prognostic value of nighttime BP for total cardiovascular end points and all-cause mortality disappeared in the subjects with perceived sleep deprivation >or=2 hours. In conclusion, nighttime BP rises and loses its prognostic significance in the hypertensive subjects who perceive a sleep deprivation by >or=2 hours during overnight monitoring.