In a previous study, rats were treated with one injection of 500 mg kg(-1) bismuth subnitrate intraperitoneally and maintained for 2 weeks. Bismuth was observed in Leydig cells, with a subsequent reduction in serum testosterone levels. In the present study, stereological procedures were used to estimate the number of Leydig cells in the right testis from the retained rats used in the previously published study. The mean number of Leydig cells in the control group was estimated to be 18.7 x 10(6) (coefficient of variation = 0.11), which is comparable to previous estimations. In the group exposed to bismuth the mean was 15.5 x 10(6) (coefficient of variation = 0.18). The observed 17% difference between the two groups was statistically significant (2P = 0.019). The inter-individual variation was largest in the bismuth-exposed group. Testis weight and body weight were not significantly reduced after bismuth exposure. These findings support the hypothesis that bismuth has a direct toxic effect on rat Leydig cells and underscores the potential risks of bismuth on male reproduction.