To determine if glipizide could enhance remission induction in new onset type 1 diabetes compared to intensive insulin treatment alone, 27 patients with type 1 diabetes were intensively treated in an open randomized trial with subcutaneous injections for one month. The insulin was randomly either discontinued (Group A) or the insulin discontinued and glipizide begun (Group B) Three patients in Group A (22%) and 7 in Group B (54%, p < .05) underwent insulin-free remissions for 10.3 +/- 4.4 and 8.7 +/- 2.6 months, respectively (p = NS). Mean blood glucose levels during insulin treatment were lower in patients entering remissions (94 +/- 3 mg/dl versus 102 +/- 5 mg/dl, p < 0.05). C-peptide levels were performed 0, 4, 8, and 24 weeks after insulin treatment. When all patients were examined, mean stimulated C-peptide levels at 4 weeks (0.58 +/- 0.09 pm/ml) were increased compared to time 0 (0.32 +/- 0.05 pm/ml, p < 0.02). Patients not entering remission had higher 4-week stimulated values (0.67 +/- 0.12 pm/ml) compared to time 0 values (0.29 +/- 0.06 pm/ml, p < .01), whereas remission patients' mean C-peptide levels remained similar at 0, 4, 8 and 24 weeks. These data indicate that a) insulin treatment plus glipizide induces higher rates of remission compared to intensive insulin treatment alone, b) the intensity of initial metabolic control may be an important determinant for remission induction, and c) endogenous insulin secretion is not associated with remission induction, suggesting that glipizide alters insulin sensitivity or is immunomodulatory in the context of new onset type 1 diabetes.