The influence of short-term thiazide treatment on peripheral tissue and liver sensitivity to insulin in insulin-dependent diabetes mellitus was determined by the euglycemic insulin clamp technique. A sequential three-step hyperinsulinemic clamp was performed in six insulin-dependent diabetics before and after 2 wk of hydroflumethiazide (HFT) administration in a daily dose of 75 mg. Insulin was infused at rates of 0.5, 2.0, and 4.0 mU X kg-1 X min-1, and each dose was given for at least 120 min. Glucose uptake during the last 30 min of each step was almost identical in the two situations (2.7 +/- 0.6 vs. 2.4 +/- 0.5 mg X kg-1 X min-1, 9.6 +/- 0.9 vs. 9.7 +/- 1.2 mg X kg-1 X min-1, and 12.0 +/- 1.3 vs. 12.6 +/- 1.5 mg X kg-1 X min-1). Serum insulin levels were also similar, and blood glucose was kept at 100 +/- 3, 99 +/- 4, and 97 +/- 3 mg/dl before thiazides and at 93 +/- 6, 93 +/- 6, and 94 +/- 6 mg/dl after thiazides. Another five insulin-dependent diabetics were infused with tritiated glucose followed by insulin infusion at two rates: 0.45 and 1.0 mU X kg-1 X min-1. Basal glucose output was comparable before and after thiazides (3.63 +/- 0.24 vs. 2.97 +/- 0.26 mg X kg-1 X min-1), as was the liver response to increasing insulin concentrations. The metabolic state as assessed by HbA1c and fasting blood glucose did not differ in the two experiments.(ABSTRACT TRUNCATED AT 250 WORDS)