In a 1-year prospective study, the outcome in infants with a platelet count less than 100 X 10(9)/L (n = 97) was compared with the outcome in an age-, weight-, and disease-matched nonthrombocytopenic control group (n = 80). The hemostatic impact of the thrombocytopenia was assessed by modified template bleeding time, hemorrhage score, and determination of the presence and extent of intraventricular hemorrhage (IVH) in thrombocytopenic infants weighing less than 1500 at birth (n = 39) compared with all nonthrombocytopenic infants less than 1500 g (n = 122) admitted during the study period. The development outcome in infants less than 1500 g was compared at 12 months after delivery. Neonatal thrombocytopenia had a major impact on hemostatic integrity: bleeding time was inversely related to platelet count (r = -0.56, P less than 0.001) and became prolonged when the platelet count fell to less than 100 X 10(9)/L. In addition, many infants (40%) had evidence of platelet dysfunction with prolonged bleeding times despite only moderately reduced platelet counts (75 to 150 X 10(9)/L). The hemorrhage score was greater in the thrombocytopenic infants compared with the sick control infants, and increased as the platelet count fell (r = -0.58, P less than 0.001). The incidence of IVH in thrombocytopenic infants less than 1500 g was 78%, compared with 48% in the nonthrombocytopenic infants (P less than 0.01). In addition, the more severe grades of IVH were more frequent in the thrombocytopenic infants. The serious neurologic morbidity for the surviving infants less than 1500 g was 41% in the thrombocytopenic infants and 7% in the nonthrombocytopenic infants. Thus, on the basis of three indices of abnormal bleeding, thrombocytopenic infants are at greater risk for bleeding than equally sick nonthrombocytopenic infants. The thrombocytopenia itself may have contributed to the high mortality and neurologic morbidity.