The purpose of this study was to determine how many days subjects should be monitored to provide an estimate of habitual physical activity in employed men engaged in a wide range of occupations. Caloric intake, movement recorders (accelerometers and pedometers), and heart rate were the measurements studied in 30 subjects who were monitored during their waking hours for 7 continuous days. A repeated measures ANOVA showed no significant difference between days (P > 0.05) for any of the measures when only weekdays were compared. However, when weekend days were included a majority of the measures showed a significant F-ratio (P < 0.05). An estimate was made of the number of days required to measure a 7-d period with less than 5% error. The average for all the different measures was 4.9 d. It appears that at least 5 or 6 d are needed to minimize the intra-individual variance a reasonable degree. Weekdays as well as weekend days need to be included.