A commercially available radioimmunoassay (RIA) kit for measurement of human adrenocorticotropin (hACTH) was validated for use in dogs. Assay sensitivity was 3 pg/ml. Intra-assay coefficient of variation (X 100; CV) for 3 canine plasma pools was 3.0 (mean +/- SD, 33 +/- 0.99 pg/ml), 4.2 (71 +/- 2.4 pg/ml) and 3.7 (145 +/- 3.7 pg/ml) %. Interassay CV for 2 plasma pools measured in 6 assays was 9.8 (37 +/- 3.6 pg/ml) and 4.4 (76 +/- 3.4 pg/ml) %, respectively. Dilutional parallelism was documented by assaying 2 pools of canine plasma at 3 dilutions and correcting the measured result for dilution. Corrected mean concentrations for the first pool were 33 (+/- 0.99), 36 (+/- 4.3), and 33 (+/- 6.8) pg/ml; corrected mean concentrations for the second pool were 145 (+/- 5.4), 141 (+/- 10.8) and 125 (+/- 3.4) pg/ml. Recovery of 1-39hACTH added to canine plasma (6.25, 12.5, 25.0, 50.0, and 100.0 pg/ml) was linear and quantitative (slope = 0.890, R2 = 0.961). To test whether anticoagulant or the protease inhibitor, aprotinin, influences ACTH concentration in canine plasma, ACTH was measured in canine blood collected in 4 tubes containing anticoagulant: heparin (H), heparin + 500 kallikrein inhibitor units (KIU) of aprotinin/ml (HA), EDTA (E), and EDTA + aprotinin (EA). Plasma ACTH concentration was the same when samples containing H and HA, or HA and E were compared, and was significantly (P < 0.01) lower in samples containing EA. Plasma storage at -20 C for 1 week or 1 month was not associated with significant change in ACTH concentration in canine plasma, using any of the 4 anticoagulant treatments. Plasma ACTH concentration measured after 6 months' storage at -20 C was significantly (P < 0.01) lower for all anticoagulants used. Synthetic 1-39hACTH added to canine blood was accurately recovered (88 to 109%, n = 3) from plasma containing EDTA, with or without aprotinin, whereas percentage recovery was overestimated by 18 to 91% in heparinized plasma. Plasma ACTH concentrations in EDTA-treated canine blood kept at 4 or 22 to 25 C for 15 to 90 minutes prior to centrifugation at 8 C were not significantly different. Plasma ACTH concentration in canine plasma was affected by storage tube material. Concentration of ACTH in canine plasma stored in borosilicate glass tubes for 1 week or 1 month at -70 C was significantly higher than initial ACTH values (P less than or equal to 0.01), but was unchanged over time in plasma stored in polypropylene or polystyrene tubes. Sample handling procedures affect canine plasma ACTH concentration measured by use of the RIA kit. Optimal sample handling conditions for plasma ACTH measurement in dogs include use of EDTA anticoagulant, blood collected at 20 to 25 C (room temperature) followed by centrifugation within 15 to 90 minutes, and plasma storage in plastic (not glass) tubes for not longer than 1 month at -20 C.