There is a need for improved reproductive toxicology assays that do not require large numbers of animals but are sensitive and informative. Therefore, Staput velocity-sedimentation separation followed by culture of specific mouse testicular cells was used as such a system. The specificity of separation was assessed using immunocytochemistry to identify spermatids, spermatocytes and spermatogonia. The efficacy of the system to detect toxicity was then evaluated by analysing the effects of hydrogen peroxide (H2O2) by the terminal uridine-deoxynucleotide end-labelling (TUNEL) assay to show the rate of apoptosis induced among the different types of germ cells. We found that 2 h of treatment at both 1¿M and 10¿M induced increases of over ~10-fold in the percentage of apoptotic cells (p¿0.001), confirming that testicular germ cells are prone to apoptosis at very low concentrations of H2O2. It was also demonstrated for the first time for this compound that spermatogonia are significantly more susceptible than spermatocytes, which are more affected than spermatids. This reflects the proportion of actively dividing cells in these cell types, suggesting a mechanism for the differential sensitivity. The approach should thus form the basis of a useful test system for reproductive and genetic toxicology in the future.