There is evidence of altered pancreatic adrenergic activity in humans with non-insulin dependent diabetes mellitus (NIDDM). We have hypothesized that altered adrenergic control of glucose metabolism is related to the pathophysiology of NIDDM, and may be evident in persons currently euglycemic, but at high risk for developing NIDDM in the future. This three year study will be conducted on the Clinical Research Unit at Duke University Medical Center. Forty women with a documented history of gestational diabetes will be recruited. Forty parous control women, matched by age and body mass index, will also be recruited. Controls will have documented normal glucose challenge in pregnancy, and no primary relative with diabetes (parent, child, sibling). Subjects will undergo oral and intravenous glucose tolerance testing after therapy with phenoxybenzamine (alpha adrenergic blocking agent) and again after placebo in a counterbalanced, single-blinded paradigm. Glucose stimulated insulin release, first phase insulin release and glucose tolerance will be assessed using a 2 X 2 analysis of variance with 2 groups and 2 conditions (drug/placebo).