The goal of this study was to assess the reliability of color Doppler imaging in boys who presented with acute scrotal pain and borderline clinical findings. The second purpose of the study was to evaluate the potential change in diagnostic accuracy as a result of employing radiological staff with varying levels of experience. Thirty-eight consecutive patients with highly suspected testicular torsion were enrolled in this prospective study. Clinical examination was followed by color Doppler US (7.5-MHz transducer). The staff of radiologists included four residents and three experienced radiologists. All patients underwent surgical exploration. All original reports of residents were reviewed by experienced radiologists and sensitivity with specificity were determined and compared. Original interpretations of color Doppler imaging yielded sensitivity of 77.8% and specificity of 85%, which changed after review to 88.9 and 90%, respectively. Clinical assessment was accurate only in 47.4%. Agreement between original and retrospective color Doppler diagnosis was obtained in 20 of 23 (86.9%) reviewed cases. Color Doppler US may prevent unnecessary surgery in the cases with conclusive normal and increased blood flow. In all other situations scrotal exploration should be performed.