The prevalence of hypertension in patients with obstructive sleep apnoea (OSA) is high and blood pressure profile is characterized by nocturnal blood pressure (BP) elevation and increased nocturnal BP variability. Ambulatory 24-hour-blood pressure monitoring (ABPM) is a valid, non-invasive method to describe circadian BP variation. Circadian BP profile and nocturnal BP variability were related to OSA severity (apnoea-hypopnoea index, mean low O2), age and body mass index (BMI) in 73 patients with OSA. Prevalence of hypertension was 75%, and in 59% BMI was greater than 30 kg m-2. A nocturnal decline of at least 10% from daytime mean BP values (night/day BP ratio <0.9; dipper) was found in only 25% of hypertensive patients and 39% of normotensive patients. Comparison between dippers and non-dippers showed significant differences in apnoea severity (apnoea-hypopnoea index 32 + 19 vs. 50 + 23/h, P < 0.01; mean low O2 84.5 + 4 vs. 80.2 + 5.8%, P < 0.01) but not for age and BMI. In multiple regression analyses with age, body mass index, apnoea-hypopnoea index and mean low O2 as independent and BP ratios and BP variability as dependent variables, sleep apnoea severity was the only independent predictor for circadian BP rhythm and nocturnal BP variability. The results presented here suggest that independent of age and obesity the severity of sleep apnoea is an important determinant of circadian BP variation and nocturnal BP variability.