To compare the minute-by-minute heart rate (Flex-HR) reference method with the factorial method, total daily energy expenditure (TDEE) and the pattern of daily energy expenditure (EE) were measured in nonpregnant, nonlactating women 19-40 yr of age, working at their household chores (at home, N = 20) or for remuneration in various kinds of employment (at work, N = 28). The factorial method used three data sets described in the compendium of Ainsworth et al. (CMD), FAO/WHO/UNU (FAO), and James and Schofield (J&S). Measurements were repeated on three rounds separated by 3 months. The TDEE by Flex-HR method, by round, was 9.0 +/- 2.6, 8.9 +/- 1.4, and 10.3 +/- 3.4 MJ.d-1 in the women at home and 9.7 +/- 2.3, 11.4 +/- 3.2, and 11.3 +/- 3.6 MJ.d-1 in the women at work. Values using the CMD data set were not significantly different from Flex-HR in either group in any round, but FAO and J&S data sets gave significantly lower values than both of the former methods in all rounds. The same statistical results were obtained for the patterns of activity. Since the timing of activities by the observers was the same, it is concluded that the major source of error in applying the factorial method is in the values for energy expenditure assigned to the various activities and the ability of investigators to properly judge EE from available data sets.