Blood pressure and heart period variability have been measured directly in 142 subjects with mild hypertension over 24 h. The variabilities have been expressed as the standard deviation of 2 min averages of all beats over 24 h. Baroreflex sensitivity was assessed in 102 subjects by the phenylephrine method. Blood pressure varies over a range of approximately 40% around the mean by day and by approximately 20% at night. The variability of blood pressure by day was inversely proportional to the sensitivity of the baroreflex (r = -0.33, P less than .001), while the variability of heart period was directly related to the sensitivity of the reflex (r = 0.27, P less than .01). Neither of these relationships was significant at night. An inverse relationship between heart period and blood pressure was shown by regression analysis of blood pressure and heart period averages over 24 h. The steepness of the slope of the heart period-systolic blood pressure relationship was strongly correlated with the baroreflex sensitivity (r = -0.55, P less than .001), suggesting that blood pressure variations are substantially buffered by changes in heart frequency. Thus, a more stable heart rate that results from an ineffective baroreflex is associated with a more variable systolic blood pressure.