Red cell volume was measured in 259 infants admitted to a high-risk newborn unit. Red cell volume was measured using 50Cr tagging which subsequently was activated to 51Cr for counting. Total blood volume was calculated using the corrected whole body hematocrit. A low red cell volume was frequently associated with a maternal history of vaginal spotting, with placenta previa or placenta abruptio, with nonelective cesarean section, and with deliveries associated with cord compression. Asphyxiated infants without a history suggestive of blood loos often had a low red cell volume. An early central hematocrit below 45% correlated with a low red cell volume, but a normal or high hematocrit was often associated with low red cell measurements. The total blood volume depended largely on whether the tagging was done long enough after the blood loss for plasma volume equilibration to have occurred. Very low red cell volume values were associated with a high mortality rate, but birth weight played a dominant role in survival, even at low red cell volume levels.