The results of duplex sonography of the cavernosal artery were compared with the results of pharmacoarteriography in a series of impotent men in order to assess the validity of sonography as a screening tool for the diagnosis of arteriogenic impotence. Duplex sonography was performed in 30 impotent men after intracavernosal injection of papaverine and phentolamine. Maximal acceleration, peak systolic velocity, and resistive index were determined for the cavernosal artery. All patients had selective pharmacoarteriography and cavernosometry, which were used to diagnose arterial disease (bilateral hemodynamically significant stenoses or occlusions), suspected arteriolar dysfunction (inconspicuous helicine arterioles), or venoocclusive insufficiency (pharmacologic maintenance erectile flow greater than 25 ml/min). Seventeen patients had arterial insufficiency, four had arteriolar dysfunction, and nine had no significant arterial disease. Significant differences in Doppler measurements were found between patients without significant arterial disease and those with arterial insufficiency, including peak systolic velocity (50 +/- 25 cm/sec vs 33 +/- 17 cm/sec, p = .04), acceleration (524 +/- 261 cm/sec2 vs 199 +/- 111 cm/sec2, p < .001), and resistive index (88 +/- 12% vs 73 +/- 11%, p = .002). Differences were also noted between the group without significant arterial disease and the group with arteriolar dysfunction. When peak systolic velocity of less than 25 cm/sec or acceleration less than 400 cm/sec2 was used as an indication of inadequate arterial patency, the sensitivities were 35% and 100%, the specificities were 61% and 46%, and the negative predictive values were 42% and 100%, respectively, in the diagnosis of arterial insufficiency. Duplex sonography of the cavernosal arteries may be a useful screening tool in patients with suspected arteriogenic impotence only when acceleration is evaluated in addition to peak systolic velocity. The specificity of the method may be partially limited by the inability to distinguish between arterial and arteriolar disease.