We evaluated the effects of gliclazide on the secretion and action of insulin in 18 non-obese (BMI 24.09 +/- 0.47 Kg/m2, range 20.89-27.52 Kg/m2) non-insulin dependent diabetic patients (mean age 56.9 +/- 1.8 years, range 39-67 years) by the oral glucose tolerance test (OGTT) and insulin suppression test (IST). Most of them were diagnosed recently and thus untreated previously. All subjects were treated with gliclazide or placebo for 3 months in a double blind cross-over design. After gliclazide therapy, fasting and 2 hour post-OGTT plasma glucose significantly decreased (136 vs. 185 mg/dl, P less than 0.005 291 vs. 358 mg/dl, P less than 0.005), 2 hours post-OGTT plasma insulin was significantly increased (79 vs. 59 uU/ml, P less than 0.05) while fasting plasma insulin remained unchanged (21 vs. 19 uU/ml, P greater than 0.1). HbAlc decreased significantly with gliclazide therapy (6.6 vs. 7.6%, P less than 0.005). In addition, oral glucose tolerance, as measured by the mean incremental areas under the plasma glucose curve, were improved significantly (1430.6 +/- 682.4 vs. 16192.5 +/- 608.5 mg.min/dl, P less than 0.005). The mean incremental areas under the plasma insulin curve during OGTT also increased (4482.0 +/- 637.1 vs. 3167.5 +/- 511.9 uU.min/ml, P less than 0.05) after gliclazide therapy. Mean steady state plasma glucose levels (SSPG) showed no remarkable difference during drug and placebo administration (188 vs. 190 mg/dl). Throughout the trial there was no significant change of body weight, nor any side effect as judged from blood counts, urinalysis, and the results of SMA-23 and EKG.(ABSTRACT TRUNCATED AT 250 WORDS)