Multiple endpoints of spermatotoxicity in short duration tests (1-5 days exposure; 2.5-week assay interval) were investigated in a number of chemicals reported to produce minimal to severe reproductive effects when administered subchronically. Six of these chemicals (boric acid, dinoseb, 2,5-hexanedione, methoxychlor, metronidazole, ornidazole) produced substantial spermatotoxicity after 1 to 5 doses. Spermatotoxic effects of chlordimeform were equivocal while p,p'-DDT, n-hexane, and sodium chlorite were judged negative. Four chemicals with known acute effects (benomyl, busulfan, ethylene glycol monomethyl ether, nitrobenzene) elicited expected histopathologic responses after a single dose. Testicular histology, testicular sperm head counts, cauda sperm counts, sperm morphology, and sperm velocity proved to be the most toxicologically sensitive endpoints in one or more of the studies, but histopathology of the testis and epididymis was the most consistent indicator of reproductive damage. The percentage of motile sperm and sperm concentration in the epididymal fluid were the least sensitive measurements. The data suggested that most chemicals with the potential to produce moderate to severe sperm damage are detectable with a short duration test. Complementary multiple endpoints enhanced the interpretation of results, often identified cellular targets, and provided insight on possible mechanisms. Specific responses were often similar to specific effects reported for subchronic exposures. A short duration test could be of value as a screen in structure-activity studies or to set priorities for chemicals requiring further evaluation. As a supplement to breeding studies, the data generated in the short test could also be used to enhance the design and interpretation of the longer tests.