Twenty-nine dogs were diagnosed with hyperadrenocorticism (HAC). A single determination of endogenous plasma adrenocorticotropic hormone (ACTH) and adrenal ultrasonography were used in a prospective study to differentiate between pituitary-dependent HAC (PDH) and adrenal-dependent HAC (ADH). In 27 out of the 29 dogs (93 per cent), both endogenous plasma ACTH concentrations and adrenal ultrasonography indicated the same cause of HAC. Twenty-one of the 29 cases (72 per cent) were shown to be pituitary-dependent; all had plasma ACTH concentrations of greater than 28 pg/ml (reference range 13 to 46 pg/ml) and both adrenal glands were ultrasonographically of similar size and of normal shape. All 21 cases responded well to mitotane therapy. Six cases (21 per cent) were shown to be adrenal-dependent; all had plasma ACTH concentrations below the limit of the assay (<5 pg/ml) and the presence of an adrenal mass on ultrasonography. The sensitivity and specificity of adrenal ultrasonography and endogenous ACTH determinations to identify the cause of HAC were demonstrated to be 100 per cent and 95 per cent, respectively, for ADH. These discriminatory tests are more accurate than published figures for dexamethasone suppression testing.