A low resting metabolic rate (RMR) for a given body composition has been identified as a risk factor for weight gain and obesity, and has also been reported in formerly obese individuals with the genetic predisposition for obesity. The possible role of thyroid hormone in low RMR was studied in a large sample of postobese women. RMR was measured by indirect calorimetry in 28 weight-stable postobese women with a family history of obesity (PO group) and in a control group of 28 nonobese women closely matched for age, fat mass, and fat-free mass. RMR was 8% lower in the PO than in the control group [[symbol: see text]; 95% Cl:5856 (5520, 6214) compared with 6408 kJ/d (6096, 6768 kJ/d), P < 0.02], and the group difference remained unchanged after fat-free mass and fat mass were adjusted for (552 kJ/d, P < 0.015). The PO group had lower plasma free triiodothyronine [2.4 (1.9, 3.0) compared with 3.4 pmol/L (2.9, 3.9 pmol/L), P < 0.01], whereas plasma androstenedione only tended to be lower in the PO than in the control group. Adjustment for differences in androstenedione did not reduce the difference in RMR, whereas adjustment for differences in plasma free triiodothyronine eliminated the group difference (96 kJ/d, P = 0.59). The present study shows that RMR for a given body composition is lower among postobese than among matched never-obese control subjects. Statistically, the lower plasma free triiodothyronine concentrations of the postobese subjects could explain their lower RMRs, but it remains to be established whether these findings are causally related.