To determine the effect of diabetes mellitus on ketone removal rates, Na DL-beta-hydroxybutyrate was administered as a continuous three-hour infusion (3 mg./kg./min.) to healthy volunteers and insulin-dependent diabetics in the postabsorptive state. An additional group of healthy controls received intravenous glucose (50 gm.) or glucose and insulin during the ketone infusion to determine the effect of hyperinsulinemia on ketone removal. Following ketone infusion, total blood ketone levels in the diabetics were twofold greater than in controls (p less than 0.001). The metabolic clearance rate of ketones (MCRk) in the diabetics was reduced by 42% from that of controls (p less than 0.001). In contrast, the calculated production rate of ketones (PRk) in diabetics was not consistently different from that observed in controls. In diabetics with normal PRk, MCRk remained significantly below control values (p less than 0.001). The ketone infusion resulted in a fall in plasma glucose and alanine levels in the normals as well as diabetics. However, the decline in plasma glucose induced by the ketone infusion was five- to sixfold greater in the diabetics than in controls (p less than 0.005) and correlated linearly with the decline in plasma alanine (p less than 0.02). Administration of intravenous glucose during an ongoing ketone infusion in normal subjects resulted in 37 +/- 5% reduction in beta-hydroxybutyrate, but no change in acetoacetate concentration. The decline in beta-hydroxybutyrate was two- to threefold greater than would be expected if glucose had acted solely to inhibit endogenous ketone production. Similar results were observed when hyperinsulinemia without hyperglycemia was produced by simultaneous administration of insulin and glucose. It is concluded that (1) ketone disposal is reduced in diabetes even when ketone production is normal, suggesting the rate of ketone utilization may be a more sensitive index of insulin deficiency than is ketone production; (2) hyperinsulinemia stimulates beta-hydroxybutyrate utilization without influencing acetoacetate concentration; and (3) increased blood ketone levels induced by infusion of Na DL-beta-hydroxybutyrate reduce plasma glucose and alanine concentrations in diabetes. These findings thus support a role for insulin in influencing ketone disposal in normal as well as diabetic man and a role for ketones in influencing substrate availability for gluconeogenesis in diabetes.