Female C3HeB/FeJ mice were fed diethylstilboestrol (DES) at a dietary concentration of 100 ppb beginning at 6 wk of age. Several oestrogen-sensitive target organs (uterus, vagina, bone and mammary gland) were removed and examined histologically when mice died, became moribund, were killed at scheduled intervals or were removed because of presumptive mammary tumours (palpable mass having a 1-cm diameter). Oestrogen-sensitive histological observations for which there was less than a 100% prevalence and/or severity included uterine fibrosis and glandular hyperplasia, vaginal mucoid stroma and keratinization, mammary adenocarcinomas and osseous trabecular proliferation or marrow fibrosis. Mice from this treated population were grouped according to the specific grade of uterine glandular hyperplasia. Each mouse having a given grade of hyperplasia was matched with a mouse that failed to develop glandular hyperplasia and was fed the 100 ppb DES diet for approximately the same time period. This procedure provided three sets of matched pairs (no hyperplasia v. grade 0; no hyperplasia v. grade 1 and no hyperplasia v. grade 2). Pairwise comparisons were then made in these groups for each of the other oestrogen endpoints. Within each set of matched pairs, mice with glandular hyperplasia had a greater prevalence and/or severity of uterine fibrosis and vaginal mucoid stroma and vaginal keratinization than those without hyperplasia and differences within matched sets were greater as the severity of hyperplasia increased. In contrast, the prevalence of osseous trabecular proliferation or marrow fibrosis were equivalent across all grades of glandular hyperplasia and the prevalence of mammary adenocarcinomas was inversely related to the severity of glandular hyperplasia. These data suggest the existence of a subpopulation of inbred mice that is especially sensitive to DES exposure for uterine and vaginal endpoints and highly insensitive to mammary tumour induction.