Perinatal survival after fetal transfusion in Winnipeg from February 1978 to June 1980 (52%—11 of 21 transfused) was worse than in the preceding 12-year period (70%—79 of 113 transfused). The cause was determined to be narrowing of the epidural transfusion catheter side hole opening diameters, which caused donor red cell hemolysis and hydrops fetalis. Catheter-induced red cell hemolysis was directly responsible for three perinatal deaths in this interval and probably contributed to two others. Catheter-induced red cell hemolysis was prevented completely by removal of the catheter tip and side hole openings, allowing donor red cell egress through the open end of the catheter. Following the institution of real-time ultrasound scan surveillance during and after intrauterine transfusion, survival for the interval from July 1980 to June 1982 was 92% (22 of 24 transfused), by far the series' best intrauterine transfusion survival rate. Hydropic fetal survival rate in the same period was 75% (six of eight transfused). With meticulous prenatal care, amniotic fluid ΔOD450 measurements beginning at 20.5 weeks' gestation, and intrauterine transfusion carried out under ultrasound guidance, beginning as early as 22.5 weeks' gestation if necessary, the Rh Laboratory has achieved extremely satisfactory perinatal salvage following intrauterine transfusion. Intensive plasma exchange, as an adjunct to the above measures, should be reserved for the pregnant woman with a history of hydropic fetal death before 28 weeks' gestation.