The effects of elevated EPI and CORT levels on KG, SI and SG were studied in dogs with alloxan-induced diabetes. Conscious dogs received SAL, EPI 20 ng μ kg−1 · min−1 for 30 min (short EPI) or 72 h (long EPI), or CORT 200 μg · kg−1 · min−1 for 60 min (short CORT) or 72 h (long CORT) before assessment of glucose metabolism by rapid sampling for glucose and insulin levels after 300 mg/kg i.v. glucose and exogenous insulin infusion designed to simulate the normal secretory pattern. With EPI infusion, KQ fell acutely from 2.9 ± 0.4 to 2.0 ± 0.2%/min (SAL vs. short EPI, P < 0.05), but rose to 3.4 ± 0.4%/min during long EPI. Minimal-model analysis of the glucose response with the insulin data as input showed that S1 decreased acutely from 4.7 ± 1.8 to 2.5 ± 0.6 × 10−5 min−1/pM (SAL vs. short EPI, P < 0.05), but rose to 4.5 ± 2.5 × 10−1 min−1/pM during long EPI. The effects of EPI on SG paralleled the results for KG and S1, with acute decline from 3.9 ± 0.4 to 2.1 ± 0.4 × 10−2 min−1 (SAL vs. short EPI, P < 0.05) and recovery to 3.3 ± 0.3 × 10−2 min−1 during long EPI. During CORT infusion, Kg tended to fall (SAL 2.9 ± 0.4 vs. short CORT 2.5 ± 0.5 vs. long CORT 2.2 ± 0.5%/min). This decline was related to a fall of S1 (SAL 4.7 ±1.8 vs. short CORT 2.7 ± 1.8 vs. long CORT 1.2 ± 0.7 × 10−1 min−1/pM, P < 0.05 long CORT vs. SAL), whereas SG levels were similar for the three groups. These results indicate that, in the absence of any compensatory change of insulin secretion, adaptation to the metabolic effects of long-term hormone elevation occurs for EPI but not CORT, which has a sustained effect on S1. Therefore, CORT may be more important than EPI as a contributor to long-term stress-induced hyperglycemia in people with type I diabetes.