Although fetal renal anomalies are frequently detected by prenatal ultrasonography, pre- and postnatal follow up of these disorders is not standardized. The investigators determined the incidence of the most common renal abnormalities seen in their ultrasound unit and correlated the presence of fetal renal pelvic dilatation (RPD) with the need for postnatal surgery in 342 prospectively studied fetuses who had isolated renal abnormalities and were followed from 3 months to 10 years. RPD was defined as an anteroposterior renal pelvis measurement of 4 mm or more at less than 33 weeks gestation or 7 mm or more afterward. Two thirds (227 of 342) of fetuses had RPD as their major sonographic finding. Obstructive lesions requiring surgical treatment were present in 60 fetuses (26% of all RPD cases), and the remainder had nonobstructive lesions that either resolved spontaneously or remained stable and did not require surgery. Approximately 70% of fetuses with obstructive and 77% with nonobstructive lesions were male. Other relatively frequent abnormalities included multicystic dysplastic kidney (10.5%), urinary tract duplication malformations (9.4%), and reflux (8.2%). The other category included bladder outlet obstruction, prune belly sequence, pelvic or horseshoe kidney, and bladder extrophy. Average RPD values were 11 mm in fetuses with nonobstructive conditions and 23 mm in those with obstruction. Ninety-five percent of fetuses with obstructive RPD underwent voiding cystourethrography (VCUG) and 92% had renal scintigraphy. VCUG demonstrated vesicoureteral reflux in 23% of these cases, and the renal perfusion scan showed urinary stasis in 96%. VCUG also was performed in 74% of fetuses with nonobstructive RPD but demonstrated reflux in only 4%. Among the 29% of nonobstructed fetuses who had a renal perfusion scan, 49% had urinary stasis. An obstructive process was identified in 10% of fetuses with a RPD of 6 to 10 mm (no fetuses with maximum RPD less than 6 mm required postnatal surgery); these fetuses accounted for approximately one fifth of the postnatal surgeries. Obstruction was also found in 27% of fetuses with a RPD of 11 to 20 mm (33% of all surgeries) and nearly 70% with a RPD exceeding 20 mm (48% of all surgeries). The overall ratio of obstructive to nonobstructive disorders was approximately 1:3. These findings endorse a conservative approach to prenatal sonographic monitoring of the kidneys. The RPD reflects the likelihood of obstruction, but there is substantial overlap, and in this ultrasound unit, RPD did not alter the timing of delivery. If RPD is detected early in pregnancy, a single follow-up sonographic study late in pregnancy or after delivery is sufficient.