OBJECTIVE—Epidemiological studies have implicated increased plasminogen-activated inhibitor 1 (PAI-1) as a marker or predictor of accelerated coronary atherosclerotic disease in type 2 diabetes. We sought to determine whether metabolic control, independent of its oral mode of implementation, affects PAI-1 in patients with marked hyperglycemia. RESEARCH DESIGN AND METHODS—A total of 91 subjects were screened, subjected to a 4-week drug washout, and randomized to daily treatment with glipizide GITS (maximum 20 mg, n = 46) or metformin (maximum 2,550 mg, n = 45) as monotherapy. After monotherapy, combination therapy was initiated by adding the second agent to the regimen. Plasma glucose (fasting and postprandial), HbA1c, fructosamine, and PAI-1 were assayed before and after randomization and sequentially thereafter in all subjects; hepatic glucose output (HGO) and abdominal fat distribution were each measured in a subset of subjects. RESULTS—Glycemic control was markedly impaired at baseline (mean HbA1c 10.4 ± 0.2% glipizide GITS; 10.0 ± 0.2% metformin) but improved comparably with each agent as monotherapy and in combination (P < 0.0001 vs. baseline), as assessed with meal tolerance studies, fructosamine values, and HGO. Body weight and abdominal fat distribution did not change significantly in either group. PAI-1 concentrations were extraordinarily high (5- to 10-fold more than normal) at baseline (202 ± 12 ng/ml glipizide GITS; 201 ± 13 ng/ml metformin) but declined comparably, and significantly, after treatment with either agent as monotherapy and decreased further with combination therapy. CONCLUSIONS—When hyperglycemia is profound, increases in PAI-1 are also profound. Control of hyperglycemia with either glipizide GITS, an insulin secretagogue, or metformin as monotherapy comparably ameliorates elevated PAI-1.