Levator ani and coccygeus muscle estrogen and androgen receptors were measured in 6, healthy, greater than or equal to 5-year-old, noncastrated, male Beagles (controls) and in 24 dogs with perineal hernia. Estrogen and androgen receptor analyses were performed on levator ani and coccygeus muscle specimens obtained from control dogs at the time of castration; contralateral levator ani and coccygeus muscle specimens were assayed 2 months after castration. During herniorrhaphy of dogs with perineal hernia, levator ani (noncastrated, n = 12; castrated, n = 7) and/or coccygeus (noncastrated, n = 5; castrated, n = 4) muscle biopsy specimens were obtained for estrogen and androgen receptor analyses. For estrogen and androgen receptor assays, each muscle biopsy specimen was homogenized in Tris-EDTA-glycerol buffer, and centrifuged at 30,000 X g; extracts were used for binding with ligands: [3H]methyltrienolone (3H-R1881) for androgen receptors, and [3H]estradiol-17 beta for estrogen receptors. Extracts were incubated overnight at 0 to 4 C. Nonspecific binding was estimated, using 100-fold concentration of cold ligands. Bound and free hormones were separated, using hydroxylapatite batch assay. Receptor numbers for each tissue were calculated as femtomoles (fmol) per milligram of protein. Quantified data were compared between precastration and postcastration controls, using a paired t-test. One-way ANOVA and Bonferroni post-hoc test were used to compare values for precastration controls, postcastration controls, castrated dogs with perineal hernia, and noncastrated dogs with perineal hernia. Significance was set at P < 0.05. Estrogen receptors were not detected. Androgen receptors were characterized by Scatchard analysis (dissociation constant = 3.16 to 6.6 nM R1881, receptor number = 23 to 175 fmol/mg of protein). Post castration controls had significantly higher numbers of androgen receptors in levator ani and coccygeus muscles than did precastration controls. Dogs with perineal hernia (castrated and noncastrated) had lower numbers of androgen receptors than did either control group. The paucity of androgen receptors in pelvic diaphragm muscles of dogs with perineal hernia, compared with controls, suggests that decreases in quantity of androgen receptors contribute to the etiopathogenesis of perineal hernia in dogs.