Several prospective studies have demonstrated a higher cardiovascular complication rate in patients with a non-dipping compared with a dipping blood pressure profile. However, the extent of blood pressure reduction during sleep can be highly variable in an individual patient, and a repeat ambulatory blood pressure study commonly shows a change in category. The reasons for this variability are not clear. In general, hypertensive patients with a non-dipper blood pressure profile have higher actigraphy-measured activity during time in bed than dipper hypertensive patients but no analysis of actigraphy-determined sleep parameters has been published. We therefore prospectively studied 52 hypertensive patients who were off antihypertensive drugs for at least 3 weeks and who underwent simultaneous ambulatory and wrist actigraphy monitoring. All patients were clinically free of any sleep-related disorder. The blood pressure profile was labelled dipper when the change in mean awake blood pressure to sleep blood pressure was 10% or higher and non-dipper if less than 10%. Overall activity levels were higher during the time in bed in the non-dipper group compared with the dipper group. An inverse relationship of daytime activity with age was seen in men (r = -0.54, P = 0.001) but not in women (r = -0.06, P = 0.79). Both groups of hypertensive patients had a similar total time in bed (458 +/- 92 min, in dippers and 433 +/- 106 min in non-dippers, P = 0.39). Sleep latency (time to onset of sleep after getting into bed) was a median of 7 min in dippers and 15 min in non-dippers (P = 0.10). In addition, dipper hypertensives spent more of the time in bed asleep than non-dippers (87% vs 76%, P = 0.004). Logistic regression analysis revealed that body mass index, kg/m(2) (OR, 1.3), and night-time activity (units/min) (OR, 1.1) were predictive of a non-dipping blood pressure profile. These data show a possible disturbance of sleep in non-dipper hypertensive patients.