Participation in at least 30 min of moderate intensity activity on most days is assumed to confer health benefits. This study accordingly determined whether the more vigorous household and garden tasks (sweeping, window cleaning, vacuuming and lawn mowing) are performed by middle-aged men at a moderate intensity of 3-6 metabolic equivalents (METs) in the laboratory and at home. Measured energy expenditure during self-perceived moderate-paced walking was used as a marker of exercise intensity. Energy expenditure was also predicted via indirect methods. Thirty-six males [ X (SD): 40.0 (3.3) years; 179.5 (6.9) cm; 83.4 (14.0) kg] were measured for resting metabolic rate (RMR) and oxygen consumption ( VO(2)) during the five activities using the Douglas bag method. Heart rate, respiratory frequency, CSA (Computer Science Applications) movement counts, Borg scale ratings of perceived exertion and Quetelet's index were also recorded as potential predictors of exercise intensity. Except for vacuuming in the laboratory, which was not significantly different from 3.0 METs ( P=0.98), the MET means in the laboratory and home were all significantly greater than 3.0 ( P</=0.006). The sweeping and vacuuming MET means were significantly higher ( P<0.001) at home than in the laboratory, whereas the converse applied for window cleaning and lawn mowing. Measured RMR was significantly lower ( P<0.001) than the 1-MET constant. Estimating METs by fitting random intercept regression models to the data resulted in standard deviations for the "leave-one-out" prediction errors (predicted-measured) of 0.4 and 0.5 METs for the laboratory and home equations, respectively. While the means indicate that all the activities were performed at a moderate intensity, there was great inter-individual variability in energy expenditure. The laboratory and home-based equations predicted with correct classification rates of 89% and 88%, respectively, whether energy expenditure was <3.0 or >/=3.0 METs.