Objective To compare the blood glucose level and associated hypoglycemia risks by using insulin Glargine or human NPH both combined with Glipizide GITS in the treatment of type 2 diabetic patients. Methods Fifty-six cases with inadequate glycemia control by sulfonylurea and/or other oral agents were randomized in two groups (3∶1). In the Glarine group, 42 patients were given Clipizide GITS 5 mg every morning and injection of Glargine at bedtime daily, while 14 patients in the NPH group were given Clipizide GITS 5 mg every morning and injection of NPH at bedtime daily. The dosage of insulin was adjusted by FBG level, seeking a target of FBG＜6.7 mmol/L, and the treatment lasted for 12 weeks. The blood glucose level and incidence of hypoglycemia were observed. The daily dosages of Glargine and NPH were recorded to analyze their relations between FBG and BMI at the beginning of the trial. Results Mean of FBG and daily glucose profile were similar in the 2 groups, but the incidence of hypoglycemia in the Glargine group was significantly lower than that in the NPH group (3 cases in the Glargine group, 7.1%, 5 cases in the NPH group, 35.7%, χ2=7.0, P =0.008). Mean daily dosages of glargine at the end point were closely related to FBG and BMI at baseline. Conclusions Bedtime injection of Glargine combined with Glipizide GITs can achieve target blood glucose control and is safer than NPH. This simple “one pill-one injection” regimen may help us achieve recommended blood glucose control targets with better patients’ compliance.