Pregnant rhesus monkeys were fed a control diet of modified cow's milk or an experimental diet containing only 25% as much as protein but made isocaloric with supplemental lactose. The ad libitum consumption of the control and experimental diets provided about 2 and 0.5 g of protein/kg of body weight per day, respectively. Total volume and energy consumption of control and experimental animals were generally comparable per kilogram of body weight, with groups showing a reduction in intake of about 20% during the latter part of pregnancy. Control females gained an average of 1.3 kg during pregnancy compared to 0.02 kg by the monkeys fed the low-protein diet. One of 15 infants born to control animals died shortly after birth; eight of 16 pregnancies in animals on the low-protein diet resulted in maternal and fetal death, stillbirth, or death in the newborn period. The birth weight of full term infants from mothers fed the low-protein diet was depressed by about 15%, and fetal linear growth was affected very little, but fetal and perinatal survival were low. We conclude that growth measurements of the newborn infants were relatively insensitive indices of the severity of the maternal nutritional deficiency and of the prognosis for the infant.