OBJECTIVE: To examine the safety and overall clinical effects of normalizing the fasting plasma glucose (FPG) level with bedtime NPH insulin alone in patients with non-insulin-dependent diabetes mellitus (NIDDM) that is poorly controlled with maximal doses of sulfonylureas. RESEARCH DESIGN AND METHODS: Twelve obese male NIDDM subjects were treated for 16 weeks with bedtime insulin after a 4-week sulfonylurea washout. The insulin dosage was increased until the FPG level was normalized. The 24-h plasma glucose profiles and lipid and HbA1c levels were measured at the beginning and end of the study, and the incidence and severity of hypoglycemic episodes were closely monitored. In addition, hyperglycemic clamp studies were performed to assess insulin secretion and provide an indirect measurement of insulin sensitivity. RESULTS: FPG (14.6 ± 0.9 mmol/l at week 0) was normalized (<6.4 mmol/l) within 6 weeks (5.9 ± 0.6 mmol/l) and remained at target levels until the end of the study (4.0 ± 0.03 mmol/l at week 16, P < 0.001). The insulin dose was 80 ± 9 U/day (0.86 ± 0.10 U/kg). Improved glycemic control was confirmed by a reduction in HbA1c (10.9 ± 0.05 vs. 7.2 ± 0.2%, P < 0.001) and mean 24-h glucose (17.2 ± 0.2 vs. 7.4 ± 0.2 mmol/l, P < 0.001). The incidence of mild or moderate hypoglycemic episodes was 3.4 ± 1/patient for the entire 16-week study, and no patient experienced severe hypoglycemia. Bedtime insulin significantly improved total cholesterol, low-density lipoprotein cholesterol, very-low-density lipoprotein cholesterol, and triglyceride levels (P < 0.01). Weight gain was 2.4 ± 0.7 kg, and blood pressure was unchanged. During the hyperglycemic clamp, there was an improvement in the first phase (P < 0.001) and in the second phase (P < 0.01) of insulin secretion. There also was an increase in the rate of exogenous glucose infused (M)(P < 0.01) and in the M/C-peptide ratio (P < 0.02), suggesting enhanced insulin sensitivity. CONCLUSIONS: NPH insulin given at bedtime in amounts sufficient to achieve a normal FPG level does not cause excessive or severe hypoglycemia and does lead to good glycemic and lipid control in NIDDM. Bedtime insulin therapy also is accompanied by improved insulin secretion and insulin sensitivity. We conclude that a single dose of insulin alone at bedtime merits consideration as a therapeutic strategy in patients with poorly controlled NIDDM.