Colour Doppler imaging and cavernosometry are the optimal means for objective assessment of pharmacologically induced penile erections. 21 impotent patients were assessed by both methods independent of the information derived from the other modality. The diagnosis obtained was found to be immediately consistent in only 10 cases, 7 of which had venous leakage. 6 of the remaining 11 patients had a diagnosis of mixed arterial and venous disease in one or the other modality, but only arterial or venous disease alone in the other. The remaining 5 patients had completely inconsistent diagnoses. However, knowledge of the clinical picture and awareness of the pitfalls in the diagnostic procedures allows most of the discrepancies to be explained. Colour Doppler imaging is subject to operator error but allows selection of patients for cavernosography.