We performed a 1-year prospective study of 807 consecutive infants admitted to a regional neonatal intensive care unit to determine the frequency, natural history, mechanism(s), and cause of thrombocytopenia. Thrombocytopenia developed in 22% of the infants. The platelet count nadir usually occurred by day 4 and resolved by day 10. Possible mechanisms responsible for the thrombocytopenia were assessed by comparing mean platelet volume, platelet-associated IgG (PAIgG), and coagulation test results in those infants whose platelet count fell below 100 X 10(9)/L (n = 97) with values in age-, weight-, and disease-matched control infants without thrombocytopenia (n = 80). In some thrombocytopenic infants, 111In-labeled-platelet survival, an estimate of megakaryocyte number in bone marrow biopsy specimens obtained at autopsy, and response to platelet infusions were also assessed. The thrombocytopenia was caused by increased platelet destruction, as shown by short 111In-labeled-platelet survival (12 to 128 hours), a rising mean platelet volume during the first week of life, normal numbers of megakaryocytes, and a poorer than predicted response to platelet infusions. A potential cause for the thrombocytopenia could be found in the majority of infants: 52% had elevated levels of PAIgG, 21% had laboratory evidence of disseminated intravascular coagulation, and 12% had had exchange transfusions. In contrast, the control infants had normal coagulation assay results, and only 15% had elevated levels of PAIgG. Birth asphyxia was identified as an associated risk factor for thrombocytopenia. This study demonstrates that transient, destructive thrombocytopenia develops in a large proportion (22%) of infants admitted to a neonatal intensive care unit, and that birth asphyxia is an important risk factor.