Cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP) are important second messengers in mediating relaxation of various smooth-muscle cells. This second-messenger pathway also appears to be essential for cavernous smooth-muscle relaxation on the basis of the assumption it would be of theoretical and clinical interest to determine the functional relevance of various phosphodiesterase (PDE) isoenzymes in human cavernous smooth-muscle. This study was concentrated on characterizing PDE isoenzymes that exist in cavernous smooth muscle and evaluating the effect of selective PDE inhibitors on relaxation that is needed for the initiation of erection. Separation of PDE isoenzymes was performed using anion-exchange chromatography [diethylaminoethanol (DEAE)-Sepharose column], and a modification of the PDE-assay method proposed by Thompson and Lakey was used. The relaxation effect of PDE inhibitors was evaluated in an organ-bath study. Three different PDE isoenzymes have been shown in human cavernous smooth-muscle homogenate: cGMP-inhibited PDE (PDE III). cAMP-specific PDE (PDE IV), and cGMP-specific PDE (PDE V). All PDE inhibitors tested showed a relaxation effect on isolated human cavernous smooth-muscle, albeit with differing potency. Quazinone (a selective PDE III inhibitor) had potency at least equal to that of papaverine (a non-selective PDE inhibitor) and had a superior effect as compared with Rolipram (a selective PDE IV inhibitor) and zaprinast (a selective PDE V inhibitor). The present study provides the rationale and opens the possibility of using selective PDE inhibitors in the treatment of patients with erectile dysfunction.