Penile Doppler ultrasonography is widely accepted as an essential examination in the diagnosis of impotence. However, measurement blood flow velocity using Doppler ultrasonography may be subject to some errors. We performed color Doppler ultrasonography in 63 patients with normal penile vascular function as diagnosed using positive responses to intracavernous pharmacological stimulation. We compared the Doppler measurement results of the 126 cavernous arteries and the ultrasonic beam angles. We used a Hitachi EUB 515, a sonographic probe of 7.5 MHz, a sampling width of 0.8 mm, a sampling depth of 1 mm, and a wall motion filter was not used. Ultrasonic beam angles were 5 to 77 degrees. The mean peak systolic velocity and end diastolic velocity values were 40. 0 cm/s and 3.9 cm/sec, respectively. The peak systolic velocity and end diastolic velocity values remained stable regardless of the ultrasonic beam angles (Kruskal-Wallis test, p = 0.56, p = 0.70). However, the variance of values became greater when the ultrasonic beam angles was larger than 55 degrees in the case of peak systolic velocity (F test, p < 0.05) and 50 degrees in the case of end diastolic velocity (F test, p < 0.05), indicating a reduction in reliability. Resistance index variance was significantly higher when ultrasonic beam angle exceeds 50 degrees (F test, p < 0.05). We believe that we should accept only those cavernous artery peak systolic velocity measurements as reliable when the ultrasound beam angle is less than 55 degrees.(ABSTRACT TRUNCATED AT 250 WORDS)