There has been a low yield of primary vesicoureteric reflux (VUR) from screening the fetal urinary tract during obstetric sonography. We sought to determine whether changing the cut-off level of fetal renal pelvic diameter from 10 mm to 4 mm would improve the yield of VUR. In a prospective community-based study, a fetal renal pelvic diameter of 4 mm or more on a transverse view of the fetal renal hilum at obstetric sonography after 16 weeks' gestation was found in 426 fetuses from 9,800 consecutive pregnancies. After birth, renal sonography was performed on 386 of the 426 babies. Of the 386 babies, 264 (187 boys) had a voiding cystourethrogram (VCUG) at a mean age of 9 weeks. Primary VUR was detected in 33 (16 boys) of the 264 infants (13%), and secondary VUR in another 5 (2%). Only 5 of the 33 (15%) babies with primary VUR would have been detected if a cut-off point of 10 mm for fetal renal pelvic diameter had been used. The prevalence of reflux was similar at each cut-off level of antenatal renal pelvic diameter from 4 to 10 mm. Neither calyceal nor ureteric dilatation was helpful in differentiating those with from those without VUR. The postnatal renal sonogram did not distinguish whether reflux was present or not. More infants with primary VUR, particularly girls, were found by changing the cut-off point for fetal renal pelvic diameter from 10 mm to 4 mm, and performing a VCUG on all such infants even if the postnatal renal sonogram was normal. Of the 33 infants with primary VUR, 9 (27%, 5 boys) had an abnormal dimercaptosuccinic acid scan. Our findings support the screening of the obstetric population for a fetal renal pelvic diameter of 4 mm or more, and then investigating the infants for VUR after birth.