A pilot case-control study was done to collect data on whether susceptibility to newborn encephalopathy and neonatal seizures is influenced by the degree of maternal-fetal sharing of HLA antigens. Cases included 13 infants with moderate or severe newborn encephalopathy and seven infants with neonatal seizures but no other signs of encephalopathy. Controls were neurologically normal infants matched to cases by date of birth, sex, race, and payment status. Infants and their mothers were typed for HLA-A, -B, -DR, and -DQ antigens. The observed frequency of sharing of maternal antigens was greater than expected (ie, 0.5) for cases compared to controls at the HLA-B, -DR, and -DQ loci but not for HLA-A. The risk of neurologic problems in the neonatal period was increased 6.3 times when there was more than one match at the HLA-DR or -DQ locus. Placental abnormalities were noted at delivery only among cases, and the mean placental weight in cases was 598 g versus 695 g in controls. Further studies with sample sizes sufficiently large to statistically test this hypothesis are needed.