Dibromochloropropane, DBCP, has had a seminal role in our current understanding of how to prevent chemical risks to health. Early toxicological studies showed its special impact on the testes, and detection of its mutagenic potency was soon followed by demonstration of its carcinogenicity to animals. Its commercial use as a useful nematocide ignored, at first, these warnings from the laboratory. When production workers, first in California and then in Israel, found they were sterile as a result of their exposure, this was convincing evidence that prevention had failed. The evidence, azospermia, oligospermia, and gonadotrophin response to testicular damage, were found first in production workers. In agricultural applicators in California who used the material, decreased sperm count and increased gonadotrophin levels were found. While production in California, Texas, and Israel was halted, studies continued and so, it seems, did use. Our first Israeli study was of workers on banana crops, and we found convincing evidence of increased spontaneous abortion in their family histories. Follow-up studies among production workers in Israel showed that some recovered testicular function, but among their offspring, otherwise in good health, there was a predominance of females. Those who did not recover from azospermia were those with high levels of follicle stimulating hormone. However data for production workers did not show increased spontaneous abortion. Nor have any studies so far shown increased birth defects or increased infant mortality. Unfavorable reproductive outcomes can be due to other agents, as shown by sprayers in Dutch orchards having hypofecundity (increased interpregnancy period) and sex ratio changes; but the agents responsible have not yet been identified. These experiences have lead to the general acceptance of some implications: (1) Paternal exposures can lead to a variety of unfavorable effects on reproductive outcome, including infertility, spontaneous abortion, and altered sex ratio. (2) Patterns of effects of potent agents in production workers and in applicators or users of chemicals may differ. (3) Although human carcinogenesis has not yet been confirmed, unfavorable reproductive outcomes are a reasonable early indicator of such risk. (4) Shifts in sex ratios of populations may be a subtle sign of more serious risks. (5) Continued use of an agent such as this under circumstances in which these UROs cannot be prevented is unconscionable.