A 5-week pilot study was conducted to determine if continuous glucose monitoring could be used to improve glycemic control. A total of nine subjects with type 1 diabetes and HbA1c values greater than 8.5% completed the study. Subjects wore a continuous glucose monitor for two 1-week periods during the study. After each sensor use, changes to diet, insulin dosage and self-monitored blood glucose (SMBG) schedule were made. HbA1c decreased from 9.9% (S.D. = 1.1%) at baseline to 8.8% (S.D. = 1.0%) 5 weeks after baseline (P = 0.0006), but daily insulin usage was unchanged over the same period of time (P = 0.428). The glucose sensors performed accurately, with a median correlation of 0.92 and a mean absolute difference of 19.1% (S.D. = 9.0%). The continuous glucose profiles allowed identification of glucose patterns and excursions that helped direct changes in therapy. These treatment changes would not have been made on the basis of meter data alone and were effective in improving glucose control. Additional studies are needed to validate these findings. This pilot study highlights the potential for continuous glucose monitoring to provide the valuable information necessary to make therapy adjustments that can dramatically improve patients' glycemic control and reduce the risk of long-term complications.