Background: Physical activity is associated with health in women. Published MET-values (MET: metabolic equivalent of task) may assess physical activity and energy expenditure but tend to be too low for subjects with a high total body fat (TBF) content and therefore inappropriate for many contemporary women. The MET-value for an activity is the energy expenditure of a subject performing this activity divided by his/her resting energy expenditure, often assumed to be 4.2 kJ/kg/h. Relationships between TBF and MET have been little studied although overweight and obesity is common in women. Available data indicate that MET-values decrease during pregnancy but more studies in pregnant contemporary women are needed. Subjects and methods: Using indirect calorimetry we measured energy expenditure and assessed MET-values in women, 22 non-pregnant (BMI: 18-34) and 22 in gestational week 32 (non-pregnant BMI: 18-32) when resting, sitting, cycling (30 and 60 watts), walking (3.2 and 5.6 km/h) and running (8 km/h). Relationships between TBF and MET-values were investigated and used to predict modified MET-values. The potential of such values to improve calculations of total energy expenditure of women was investigated. Results: The resting energy expenditure was below 4.2 kJ/kg/h in both groups of women. Women in gestational week 32 had a higher resting energy metabolism (p amp;lt; 0.001) and 7-15% lower MET-values (p amp;lt; 0.05) than non-pregnant women. MET-values of all activities were correlated with TBF (p amp;lt; 0.05) in non-pregnant women and modified MET-values improved estimates of total energy expenditure in such women. In pregnant women, correlations (p amp;lt;= 0.03) between TBF and MET were found for running (8 km/h) and for walking at 5.6 km/h. Conclusions: Our results are relevant when attempts are made to modify the MET-system in contemporary pregnant and non-pregnant women. MET-values were decreased in gestational week 32, mainly due to an increased resting energy metabolism and studies describing how body composition affects the one MET-value (i.e. the resting energy metabolism in kJ/kg/h) during pregnancy are warranted. Studies of how pregnancy and TBF affect MET-values of high intensity activities are also needed. Corrections based on TBF may have a potential to improve the MET-system in non-pregnant women.