OBJECTIVE To test whether a two-injection regimen of HUL/R would improve FBG and metabolic control in pediatric IDDM patients with a dawn rise in FBG compared with our standard twice-daily therapy, HL/R. RESEARCH DESIGN AND METHODS Seventy-seven patients with fasting hyperglycemia (prebreakfast mean FBG greater than or equal to 8.3 mM (150 mg/dl) during the preceding 2 wk) were evaluated with twice-weekly midsleep (0230-0330) FBG for 2 wk. Forty-seven patients (61%) had a mean dawn rise between midsleep and prebreakfast of greater than or equal to 2.8 mM (50 mg/dl). Patients continued on HL/R for an additional 4 wk, after which 31 patients were then randomized into a double-blind 12-wk trial of either HUL/R (n = 14) or HL/R (n = 17) administered before breakfast and the evening meal. Midsleep FBG was obtained twice weekly with weekly insulin adjustment as needed to optimize glycemic control. FBG was monitored and verified with memory glucometers (Glucometer M). HbA1c levels were measured at the time of physician visits at 0, 6, and 12 wk. RESULTS Prebreakfast FBG was lower in the HUL/R-treated patients (10.6 +/- 0.6 vs. 12.6 +/- 0.6 mM [191 +/- 6.4 vs. 227 +/- 11.2 mg/dl], P less than 0.02). The dawn rise was diminished in the HUL/R patients (0.5 +/- 0.5 vs. 2.6 +/- 0.7 mM [9 +/- 8.3 vs. 46 +/- 11.7 mg/dl], P less than 0.02). FBG at lunch, dinner, bedtime, and midsleep were similar in both groups, and HbA1c did not differ between groups or change significantly in either group during the 12-wk trial. Insulin dose, percentage R, day-night dosage split, and episodes of hypoglycemia (FBG less than 3.3 mM [60 mg/dl]) were similar in both groups. CONCLUSIONS A 12-wk trial of twice-daily HUL/R improved fasting glycemia in pediatric patients with a dawn rise but did not improve metabolic control as measured by HbA1c.