Color Doppler ultrasound has been reported to be useful in ruling out testicular torsion in children. Some investigators, however, have continued to report false-positive and false-negative studies. Herein we review our institution's experience over the past 6 years. From January 1990 to December 1996, over 300 boys 21 years of age or less underwent color Doppler ultrasound for testicular pain and/or swelling. Presentation, ultrasound diagnosis, operative findings (where applicable), and subsequent course were assessed. Complete data was available for 243 boys (average age 10.36 years, range 1 day to 21 years). Fourteen boys were explored without imaging based on strong clinical findings and short duration of pain. Of these, 13 had torsion and 1 had epididymitis. Torsion was ruled out by ultrasound in 182 with 100% specificity verified on follow-up. Forty-five diagnoses of torsion by ultrasound were confirmed operatively. Two patients with equivocal scans also were explored: one had torsion, whereas the other had a torsed appendix. Ultrasonic findings are characterized. Additionally, two cases of false-negative ultrasound from outside institutions are discussed. In conclusion, color Doppler ultrasound can identify reliably those children with an acute scrotum who require exploration and exclude those children without testis torsion who would otherwise undergo needless surgery.