461 consecutive inborn babies, delivered during the period September 1993-March 1994 were subjected to a microhematocrit assessment at 6 hours of age to determine the incidence of polycythemia. 47 babies (inborn and out born), admitted to the neonatal unit with confirmed polycythemia were studied for clinical and laboratory abnormalities. These 47 babies were then randomly assigned to receive partial exchange transfusion with either normal saline or fresh plasma. The incidence of polycythemia was 27 of 461 (5.8%). 23 of 27 (85.1%) were term babies and 15 of 27 (55.5%) were small for gestational age. 14 of 27 (51.1%) babies had mothers who had pregnancy induced hypertension. Feeding problems (16/47) and lethargy (25/47) were the commonest symptoms (34% and 51% respectively). 25 of 47 (51%) babies had hypoglycemia and 5 of 47 (10.6%) had hypocalcemia. Thrombocytopenia was seen in 13/47 (27.65%) of cases. 24 babies received normal saline and 23 received fresh plasma for partial exchange transfusion. The immediate post-exchange fall in hematocrit was significant in both groups and this was well sustained over the following 48 hours. However, improvement in clinical and laboratory parameters was more remarkable with fresh plasma. Polycythemia appears to be a real clinical entity in neonates in India and babies with known risk factors should be actively screened for this condition. Once diagnosis is established special attempts should be made to rule out hypoglycemia. For treatment of polycythemia fresh plasma is preferable for partial exchange transfusion but normal saline appears to be an adequate substitute.