This randomised, open-label, two-way cross-over study compared the coefficient of variance (CV) of fasting and postprandial blood glucose (FBG and PPBG) with insulin glargine (glargine) versus neutral protamine Hagedorn (NPH) insulin treatment in patients with Type 2 diabetes (T2DM). Patients (N=20) on oral antidiabetic drugs (OADs) were treated with NPH (at bedtime) or glargine (at dinnertime) for 12 weeks of each cross-over treatment period; OADs were continued. The FBG CV was calculated from self-monitored BG values and PPBG using venous blood samples, or continuous glucose monitoring system (CGMS). Both insulins provided similar improvements in glycaemic control; however, PPBG was significantly lower after a standard meal test (performed at 13:00 h the day after insulin injection) with glargine versus NPH (p=0.02). Thirteen versus 15 patients experienced >or=1 episode of hypoglycaemia with glargine versus NPH. The results suggest that glargine plus OADs is more effective in reducing PPBG fluctuations during the day than NPH plus OADs.