Dams with 7 pups each were randomly assigned to two different diets. Twelve dams were fed a normal (20%) protein diet and were divided into two groups of 4 and 8 animals. Pups from group 1 (n = 28) were injected with citrate buffer as a control. Pups from group 2 (n = 56) were injected with streptozotocin. Twelve additional dams were fed a 40% protein diet. They were also divided into two groups of 4 and 8 animals. Pups from group 3 (n = 28) were injected with citrate buffer as a control. Pups from group 4 (n = 56) were injected with streptozotocin. Forty-eight hours later, diabetic status was determined using Dextrostix. On Day 15, pups were injected with [14C]proline to determine collagen synthesis and 45Ca to study mineralization. After the pups were killed, blood glucose levels were determined. Then mandibles were removed. Milk from each dam was also collected after injection of oxytocin. At the time of killing, blood glucose levels in diabetic pups were less than earlier levels, though still higher than those of controls on either diet. The weights of body and mandible, collagen contents, and the total calcium contents in the diabetic group were in general less than those of the nondiabetic group on the 20 and 40% protein diets. 45Ca uptake in the diabetic group was significantly increased compared with those of the nondiabetic rats on both diets. The percentage reduction in the mandibles of diabetic rats from those of nondiabetic rats on the 40% protein diets was consistently less than that of animals on the 20% protein diets. The higher protein contents of the maternal milk in the 40% protein group may partly be responsible for the smaller impairment of mandibular development in the diabetic over nondiabetic animals. It is concluded that maternal low-carbohydrate high-protein diets will play indirectly a beneficial role in the development of the mandibles of diabetic newborns.