Our purpose was to evaluate the impact of non-invasive assessment of fetal anemia and anti-D antibody quantification on the timing and frequency of invasive procedures in pregnancies complicated by rhesus alloimmunization. Nineteen consecutive non-hydropic pregnancies referred to the fetal medicine center were assigned a prior risk category (none/mild, moderate or severe) and monitored by: (1) serial fetal measurements of umbilical vein maximal flow velocity (UVVmax), liver length and spleen perimeter measurements; and (2) serial anti-D antibody concentration. Invasive tests for fetal anemia (amniocentesis or fetal blood sampling) were deferred in the absence of abnormal ultrasound findings and/or rising antibody levels. In six cases serial non-invasive tests were normal with stable antibody levels, and no invasive tests were performed; four infants were mildly affected, one was unaffected and one required postnatal exchange transfusion. In the remaining 13 affected cases, amniocentesis was performed in nine cases for: elevated UVVmax alone (n = 3), elevated UVVmax and an increased antibody level (n = 2), or normal UVVmax with an increased antibody level (> 15 IU/ml) and severe prior risk category (n = 4). Six fetuses underwent fetal blood sampling (initial hematocrit 9-29%), and five of these had an elevated UVVmax. Liver length and spleen perimeter measurements were increased in only one anemic fetus (hematocrit 13%). Of 17 infants born alive, an elevated UVVmax prior to delivery was predictive of the need for exchange transfusion (six of seven cases with an elevated UVVmax vs. one of ten with a normal UVVmax; chi 2 = 5.73, p = 0.017 with Yates' correction). These preliminary data suggest that pregnancies with a mild or no history of fetal anemia may be monitored by a combination of serial antibody quantification and Doppler monitoring of UVVmax.