78 patients (pts.) with erectile dysfunction (ED) underwent conventional diagnostic evaluation including NPTR monitoring with the Rigiscan, intracavernous drug testing, dynamic cavernosometry, doppler evaluation, pharmaco-angiography and neurologic tests. Only erections of the best rigidity recorded over three nights served for classification into three classes of rigidity. Patients with neurogenic impotence showed significantly poorer and fewer erectile events per hour than vasculogenic impotent men. Patients with arteriogenic ED had better erections than patients with venogenic or mixed arterio-venogenic ED. Severity of organic ED during conventional diagnostic work-up correlated to loss of rigidity during NPTR recordings, but no differential diagnosis between arteriogenic, venogenic or neurogenic impotence could be made with NPTR data alone. The presented criteria for NPTR evaluation permit a time-saving, examiner-independent analysis.