Serum concentrations of 17-hydroxyprogesterone and cortisol were measured before and after the administration of exogenous adrenocorticotrophic hormone (ACTH) to three groups of dogs: 27 healthy dogs (group 1), 19 dogs with non-adrenal illness (group 2) and 46 dogs with hyperadrenocorticism (group 3). The median (range) post-ACTH concentrations of 17-hydroxyprogesterone were 5.0 (22.2 to 16.8), 6.9 (2.0 to 36.2) and 14.4 (1.7 to 71) nmol/litre in groups 1, 2 and 3, respectively. There were no significant differences in the basal or post-ACTH concentrations of cortisol or 17-hydroxyprogesterone between groups 1 and 2. The post-ACTH concentrations of 17-hydroxyprogesterone in group 3 were significantly (P<0.001) greater than those in groups 1 and 2 combined. The area under the receiver operating curve (ROC) for the post-ACTH concentration of cortisol (0.94) was significantly greater than that for the post-ACTH concentration of 17-hydroxyprogesterone (0.76). Using a two-graph ROC analysis, a cut-off of 8.5 nmol/litre was found to maximise both the sensitivity and specificity of the post-ACTH concentration of 17-hydroxyprogesterone for the diagnosis of hyperadrenocorticism at 71 per cent. With a cut-off of 4.5 nmol/litre the sensitivity increased to 90 per cent but the specificity decreased to 40 per cent; with a cut-off of 16.7 nmol/litre the specificity increased to 90 per cent but the sensitivity decreased to 47 per cent.