OBJECTIVE Radiographic evaluation for prenatal hydronephrosis often includes voiding cystourethrography to ascertain whether vesicoureteral reflux is present. We sought to determine whether use of voiding cystourethrography could be limited to those patients at greatest risk for vesicoureteral reflux. We hypothesized that vesicoureteral reflux could be predicted by findings on renal/bladder ultrasonography of hydroureter, renal dysmorphia and/or duplication. METHODS We reviewed the records of patients with prenatal hydronephrosis who underwent initial postnatal ultrasonography and voiding cystourethrography during a 3-year period. The presence of vesicoureteral reflux on voiding cystourethrogram was correlated to ultrasound findings, including hydronephrosis grade, presence of hydroureter, renal dysmorphia or duplication, with ultrasound considered positive for any of the latter 3 findings. RESULTS Of 262 patients 47 (18%) had vesicoureteral reflux. Ultrasound was positive in 24 of 29 patients (83%) with high grade reflux and 12 of 18 (67%) with low grade reflux. If ultrasonography showed any of the 3 positive findings, the odds ratio of detecting vesicoureteral reflux was 8.07 (95% CI 3.86, 16.87). Using these criteria, among all cases of prenatal hydronephrosis 5 (2%) with high grade vesicoureteral reflux and 6 (2%) with low grade reflux would have been missed. Among the 47 cases of reflux overall 5 of 29 high grade (17%) and 6 of 18 low grade cases (33%) would have been missed. CONCLUSIONS By using ultrasonography criteria of hydroureter, duplication and renal dysmorphia for patients with prenatal hydronephrosis, vesicoureteral reflux can be detected more specifically. Using our criteria, 165 of 262 voiding cystourethrograms (63%) could have been avoided in patients with prenatal hydronephrosis during a 3-year period. Reducing these evaluations may decrease risks regarding radiation exposure, family anxiety and health care costs.