This study was initiated to 1) assess gemfibrozil's ability to lower plasma triglyceride (TG) concentration in patients with NIDDM, and 2) determine whether this effect was associated with any changes in glycemic control. Measurements were made of mean hourly plasma glucose, insulin, TG, and FFA concentrations from 1200-1600 h in response to a test meal; hepatic glucose production (HGP); insulin-stimulated glucose uptake during a hyperinsulinemic glucose clamp study (MCR); and fasting plasma lipoprotein TG and cholesterol concentrations in 12 patients with NIDDM before and 3 months after gemfibrozil treatment. Although ambient plasma TG and FFA concentrations fell significantly, plasma glucose, insulin, HGP, and glucose MCR did not change. However, when patients were divided into two groups, those with fasting plasma glucose levels above 9 mmol/L (fair control) and those with levels below 9 mmol/L (good control), a different phenomenon was observed. Patients in fair control had significant decreases in mean hourly plasma concentrations of glucose (15.1 +/- 1.7 to 12.6 +/- 0.9 mmol/L; P < 0.001), insulin (523 +/- 59 to 471 +/- 75 pmol/L; P < 0.001), FFA (652 +/- 150 to 504 +/- 76-mu-mol/L), and HGP (9.5 0.4 to 8.1 +/- 0.4-mu-mol/kg.min; P < 0.005), and an increase in glucose MCR (2.63 +/- 0.49 to 3.72 +/- 0.54 mL/kg.min; P < 0.07) in association with a fall in TG from 6.9 +/- 1.3 to 3.5 +/- 0.9 mmol/L (P < 0.001). Although fasting low density lipoprotein cholesterol increased (1.8 +/- 0.2 to 2.7 +/- 0.2 mmol/L; P < 0.05), the ratio of total to high density lipoprotein cholesterol decreased (6.84 +/- 0.8.8 to 5.80 +/- 1.05; P < 0.02). Despite a significant fall in mean hourly TG concentration (4.6 +/- 0.7 to 3.8 +/- 0.7 mmol/L; P < 0.001), neither insulin, FFA, HGP, nor glucose MCR changed in patients in good control. Furthermore, the mean hourly plasma glucose concentration increased from 9.2 +/- 0.7 to 11.7 +/- 1.4 mmol/L (P < 0.001). Low density lipoprotein cholesterol also increased in this group (1.9 +/- 0.2 to 2.7 +/- 0.2 mmol/L; P < 0.02), but, as before, the ratio of total to high density lipoprotein cholesterol decreased (8.15 +/- 1.93 to 6.36 +/- 1.03; P < 0.02).