Minute-by-minute heart-rate recordings over a period of 24 hours were obtained once for each of a group of 29 elderly women, 14 of whom lived in private accommodation, and 15 of whom lived in institutionalized homes for the aged. Activity diaries were kept on the subjects during the measurement period. A comparison of individual 24-hour profiles, diaries, and group averages was made. Strong positive correlations were found in both groups between average waking and sleeping heart-rates, and between the standard deviations of the two periods. A marked predominance of sedentary activities among the institutionalized women was found to correspond with lower waking heart-rates, and flatter 24-hour profiles. However, these characteristics could also be explained as an age effect. Cosinor analysis indicates that, for the majority of these subjects, a curve made up of a combination of two periodic components with maxima around midday and 8 pm affords a reasonable representation of diurnal variation.