Cardiovascular disease and diabetes may be programmed early in life by abnormal development associated with undernutrition. We investigated whether maternal nutritional status (MNS; height, pregnancy weight gain, nonpregnant BMI, and prenatal supplementation) or childhood nutritional status (CNS; birth weight, length, ponderal index, height-for-age Z-score at 24 mo, and supplementation from 0 to 24 mo) were related to fasting plasma glucose levels in rural-born Guatemalan adults. We studied 209 men and 220 women (mean age 24.4 y) who were involved in a randomized trial of nutritional supplementation of their mothers during pregnancy and during their early childhoods, conducted from 1969 to 1977. In 2 villages, residents were offered Atole (3.8 MJ and 64 g protein/L); 2 other villages were offered Fresco (1.4 MJ/L, no protein). No associations were observed between anthropometric measures of MNS or CNS and fasting plasma glucose levels. In subgroup analyses, inverse associations (all P < 0.15) with birth size were found among women born to fatter mothers, women with low supplement intake, men born to short mothers, and men more severely stunted at 24 mo. Prenatal supplementation was inversely associated with fasting plasma glucose among women [-0.40 +/- 0.17 mmol/(L. MJ. d), P = 0.02]. Among men, postnatal intake of supplementation of 0.10 to 0.20 MJ/d was associated with up to a 0.56 mmol/L reduction in fasting plasma glucose (P = 0.03), but intake in excess of 0.20 MJ/d provided no added benefit. Among women, the benefit of postnatal supplementation was restricted to those born thin (test for interaction P = 0.10). Improving the nutritional status of undernourished women and children may have positive long-term consequences.