PURPOSE: To compare the usefulness of contrast-enhanced voiding ultrasonography(US) with that of radiographic voiding cystourethrography(VCUG) for the diagnosis of vesicoureteral reflux(VUR) in children. MATERIALS AND METHODS: Ninety-five kidney-ureter units of 47 patients referred for investigation of VUR underwent contrast-enhanced voiding US followed by radiographic VCUG. After baseline US examination of the urinary tract, residual urine in the bladder was drained through an inserted Foley catheter and the bladder was gravity filled at a height of 1 m with normal saline. A galactose-based, microbubble-containing echo-enhancing agent (Levovist; Schering, Berlin, Germany) was then administered. The amount of this was approximately 10% of bladder capacity, and VUR was diagnosed when microbubbles appeared in the ureter or pelvo-calyceal system. Using radiographic VCUG as a reference point, the accuracy with which contrast-enhanced voiding US detected VUR was calculated. RESULTS: In 87 of 95 kidney-ureter units (91.6%), the two methods showed similar results regarding the diagnosis or exclusion of VUR, which was detected by both in 12 units, but by neither in 75. VUR was shown to occur in a total of 20 units, but in eight of these by one method only. In two units, VUR detected by contrast-enhanced voiding US was not demonstrated by radiographic VCUG; in six units, the reverse was true. In the detection of VUR, contrast-enhanced voiding US showed a sensitivity of 66.7%, a specificity of 97.4%, a positive predictive value of 85.7%, and a negative predictive value of 92.6%. CONCLUSION: Contrast-enhanced voiding US is highly specific and has high positive and negative predictive values; its sensitivity, however, is not sufficiently high. The modality appears to be a useful diagnostic tool for the detection of VUR without exposure to ionizing radiation, though to be certain of its value, more experience of its use is first required.