Whether fetal growth retardation could be induced by diets apparently adequate in protein and energy sources, but imbalanced relative to a single amino acid, was investigated. Pregnant rats were fed one of three diets: either a basal (B) 6% casein diet supplemented by L-methionine (L-Met) and a 5% mixture of essential plus 4.5% small neutral non-essential amino acids, which provided 4 kcal and 0.16 g protein-equivalent/g diet and was complete in other essential nutrients; an experimental (E) diet [same, except 0.4% L-threonine (L-Thr) supplement omitted]; or a control (C) diet, life B, but containing 20% casein. A total of 343 fetuses and their placentas were delivered by Caesarean section usually on day 21 of pregnancy but occasionally on day 20 or 22. Food consumption was similar in all groups (kcal/100 g rat/day); however, protein intake of the B and E rats was bout half that of the C group. Net maternal weight gain (excluding uterus + contents) was +8% in C. )% in B and -12% in E group dams. Body weight, length, volume, brain weight and placental weight, adjusted by multiple regression analysis to standardize for maternal prepregnant weight, length, litter size, net weight gain and duration of gestation, were significantly reduced in 155E greater than 156 B greater than 32C fetuses. Protein restriction accounted for 30-59% of the total observed variance in birth measures of the B versus C group fetuses; and the Thr-limiting amino acid imbalance for 36% of the total variance between the E and B groups. Imbalanced diets containing excess dispensable amino acids with adequate net protein and caloric intakes during pregnancy induce fetal growth retardation (F GR), which is increased by limiting threonine. Net maternal weight loss does not prevent F GR in rats under these conditions.