A summary of the results of the studies conducted in the EU Project "Multi-endpoint analysis of genetic damage induced by 1,3-butadiene and its major metabolites in somatic and germ cells of mice, rats and man" is presented. Results of the project are summarized on the detection of DNA and hemoglobin adducts, on the cytotoxic and clastogenic effects in somatic and germinal cells of mice and rats, on the induction of somatic mutations at the hprt locus of experimental rodents and occupationally exposed workers, on the induction of dominant lethal mutations in mice and rats, and on heritable translocations induced in mice, after exposure to butadiene (BD) or its major metabolites, butadiene monoepoxide (BMO), diepoxybutane (DEB) and butadiene diolepoxide (BDE). The primary goal of this project was to collect experimental data on the genetic effects of BD in order to estimate the germ cell genetic risk to humans of exposure to BD. To achieve this, the butadiene exposure are based on data for heritable translocations and bone marrow micronuclei induced in mice and chromosome aberrations observed in lymphocytes of exposed workers. A doubling dose for heritable translocations in human germ cells of 4900 ppm/h is estimated, which, assuming cumulative BD exposure over the sensitive period of spermatogenesis, corresponds to 5-6 weeks of continuous exposure at the workplace to 20-25 ppm. Alternatively, the rate of heritable translocation induction per ppm/h of BD exposure is estimated to be approximately 0.8 per million live born, compared to a spontaneous incidence of balanced translocations in humans of approximately 800 per million live born. These estimates have large confidence intervals and are only intended to indicate orders of magnitude of human genetic risk. These risk estimates are based on data from germ cells of BD-exposed male mice. The demonstration that clastogenic damage was induced by DEB in preovulatory oocytes at doses which were not ovotoxic implies that additional studies on the response of mammalian female germ cells to BD and its metabolites are needed. The basic assumption of the above genetic risk estimates is that experimental mouse data obtained after BD exposure can be extrapolated to humans. Several points exist in the present report and in the literature which contradict this assumption: (1) the level of BMO-hemoglobin adducts was significantly elevated in BD-exposed workers; however, it was considerably lower than would have been predicted from comparable rat and mouse exposures; (2) the concentrations of the metabolites DEB and BMO were significantly higher in mouse than in rat blood after BD exposure. Thus, while metabolism of BD is qualitatively similar in the two species, it is quantitatively different; (3) no increase of HPRT mutations was shown in 19 workers exposed on average to 1.8 ppm of BD, while in a different population of workers from a US plant exposed on average to 3.5 ppm of BD, a significant increase of HPRT variants was detected; and (4) data from cancer bioassays and cancer epidemiology suggest that rat is a more appropriate model than mouse for human cancer risk from BD exposure. However, the dominant lethal study in rats gave a negative result. At present, we do not know which BD metabolite(s) may be responsible for the genetic effects even though the bifunctional alkylating agent DEB is the most likely candidate for the induction of clastogenic events. Unfortunately, methods to measure DEB adducts in hemoglobin or DNA are only presently being developed. Despite these several uncertainties the use of the mouse genetic data is regarded as a justifiable and conservative approach to human genetic risk estimation given the considerable heterogeneity observed in the biotransformation of BD in humans.