To observe the effect of the dawn phenomenon on basal glucose and postbreakfast hyperglycemia in patients with type 1 diabetes treated with once-nightly insulin glargine and premeal insulin lispro.In 49 study subjects consuming a fixed isocaloric (50% carbohydrate) diet of usual food, the insulin glargine dose was titrated from daily continuous glucose monitoring downloads to achieve a basal glucose goal of <130 mg/dL 4 hours after meals and during serial meal omissions but with fewer than 10% of readings at <70 mg/dL during 24 hours. Patients also performed self-monitoring of plasma glucose 7 times a day (before and 2 hours after each meal or omitted meal and at bedtime).The target mean basal glucose level was achieved only during the non-dawn phenomenon period (1400 hours to 0400 hours). During the dawn phenomenon, the mean (standard deviation) basal glucose level increased from 118 (57) mg/dL at 0400 hours to 156 (67) mg/dL before the breakfast meal, a 32% increase (P = .00149). The mean self-monitored plasma glucose level with meal omission was 63.8% of that increase with a breakfast meal.The fasting morning glucose concentration is considerably elevated because of the dawn phenomenon. Targeting insulin titration to this glucose level may result in excessive basal insulin dosing for the non-dawn phenomenon periods of the day. The dawn phenomenon is a large component of the postbreakfast hyperglycemia. Rather than increasing the morning premeal insulin bolus, consideration should be given to pretreating the earlier dawn phenomenon with an insulin pump with use of a variable basal insulin rate.