1. We have previously shown that among alpha 1-adrenoceptor antagonists used or investigated for the treatment of benign prostatic hyperplasia, tamsulosin discriminates alpha 1-adrenoceptor subtypes in rat tissues whereas alfuzosin and naftopidil do not. We now expand these studies to additional drugs (doxazosin, terazosin) being used and/or investigated for this purpose, and have evaluated all of these drugs at cloned subtypes and in human prostate. 2. Competition binding studies were performed with [3H]-prazosin in membrane samples from rat spleen, kidney and cerebral cortex and human prostate and with cloned alpha 1-adrenoceptors expressed in COS cells. Doxazosin and terazosin did not discriminate alpha 1-adrenoceptor subtypes in rat kidney and cerebral cortex. In contrast, the subtypes present in the tissues were well discriminated by the alpha 1A-adrenoceptor-selective reference drug WB 4101. 3. Alfuzosin, doxazosin, naftopidil and terazosin did not discriminate cloned alpha 1-adrenoceptor subtypes transiently expressed in COS cells whereas tamsulosin and WB 4101 did. 4. In human prostate, alfuzosin, doxazosin, naftopidil and terazosin did not discriminate the alpha 1-adrenoceptor subtypes present in this tissue whereas tamsulosin and the alpha 1A-adrenoceptor-selective reference drugs WB 4101, phentolamine and 5-methylurapidil did. Based on data with the alpha 1A-adrenoceptor-selective drugs, human prostate contains alpha 1A- and alpha 1B-adrenoceptors in an approximate 70:30% ratio. 5. We conclude that tamsulosin, in common with WB 4101, but in contrast to alfuzosin, doxazosin, naftopidil, and terazosin is selective for alpha 1A-adrenoceptors which appear to dominate in the human prostate; the therapeutic relevance of this selectivity remains to be assessed in clinical studies.