It has been suggested that the lower birth weight of newborn infants of mothers who smoke is due mainly to a deficit of lean body mass (LBM). We tested this hypothesis by measuring total body potassium, thus deriving the LBM and fat mass, of newborn infants of mothers who smoked (I-SM; n — 32) or did not smoke (INSM; n=46). Mothers who smoked were significantly younger than nonsmoking ones (25.4 and 28.9 yr, respectively) and with less years of education, but were similar in other parameters examined. The 78 infants, all singleton, were studied within 1 to 3 days of birth. The I-SM had significantly reduced birth weight, length, and head circumference but there was no difference in skinfold thickness. We measured total body potassium with a whole-body counter specially made for use with infants. Mean absolute total body potassium was significantly greater in the INSM, but the concentration in relation to weight was not different. Assuming 1 kg LBM to contain 52.1 mmol potassium, the mean LBM was 3028 g in the I-NSM and 2739 in the I-SM; mean fat mass was similar in both groups. Multiple regression analysis revealed an independent negative effect of mother's smoking on birth weight and LBM. This unequal reduction in LBM indicates a complex effect of smoking, probably mediated by alterations in protein synthesis and adipocyte metabolism. It may relate to the higher morbidity rates in infants of mothers who smoke.