To evaluate whether circadian rhythms in blood pressure and heart rate are influenced by age, we analyzed 24-h ambulatory blood pressure and heart rate recordings from 31 patients with primary hypertension. Data were collected during hospitalization, after a drug-free run-in period. Set times were administered for lights-on, meals, and lights-off. Daytime napping was prohibited. The patients were divided into sex-matched groups of young (group I: 25-45 years, n = 9), middle-aged (group II: 47-57 years, n = 11), and old (group III: 57-74 years, n = 11) subjects. Hourly data were analysed by fitting a two-component cosine function (24- and 12-h periods). Amplitudes of the circadian rhythms in systolic blood pressure and heart rate were significantly reduced with age. This finding could be partly attributed to the recording of higher nocturnal values in older patients. Elderly hypertensives also evidenced a significantly greater ultradian component (12-h period) in the systolic blood pressure rhythm than did young patients, with the secondary afternoon decline in blood pressure being more pronounced in groups II and III. The 24-h acrophase of heart rate was found to occur 1.6 h earlier than that of systolic blood pressure in the young group (p < 0.01). This phase advance of heart rate compared with systolic blood pressure was reduced to 1 h in group II (p < 0.05) and was not evident in group III (p > 0.1). These results indicate that circadian blood pressure and heart rate profiles of primary hypertensives change with age. Since measures were obtained in a typical clinical setting, these findings have implications for the diagnosis and treatment of hypertension in the elderly. The marked afternoon decline in blood pressure for the elderly patients may also render conventional cosinor analysis inappropriate for accurate description of the circadian rhythms of geriatric hypertensives.