Basal plasma glucose is usually increased in uncomplicated malaria, implying insulin resistance. If the infection progresses, the risk of hypoglycemia will increase as host glucose production becomes insufficient for host/parasite demand. To assess the relative contribution of insulin-mediated and non-insulin-mediated glucose disposal to plasma glucose levels in severe malaria, we studied six healthy controls (two males and four females; mean age, 38 years) and eight patients with complicated falciparum malaria (five males and three females; mean age, 31 years) who had a frequently sampled intravenous glucose tolerance test (FSIVGTT) in which 10% of the dextrose bolus was 6,6-D2-glucose. The minimal model was applied to native and labeled plasma glucose and serum insulin profiles over 4 hours postinjection. Basal plasma glucose concentrations in the patients were significantly greater than in the controls (median [range], 6.1 [2.1 to 8.5] v 4.3 [3.9 to 4.7] mmol/L, P = .03). Malaria-associated insulin resistance was confirmed by a lower insulin sensitivity index (SI) in patients (5.6 [2.4 to 17.4] v 16.0 [2.5 to 22.3] x 10(-4).min-1 per microU/mL in controls, P = .026). Glucose effectiveness ([SG] the ability of glucose to reduce its own plasma concentration) was higher in the patients (0.015 [0.006 to 0.024] v 0.008 [0.007 to 0.010] min-1 in controls, p = .019). Glucose disappearance at basal concentration was increased by a median of 42% in severe malaria patients, with the insulin-independent component comprising 81%, versus 67% in controls. Indices of beta-cell function were normal in malaria patients. These data demonstrate that basal plasma glucose utilization is increased approximately 50% in severe malaria, consistent with previously published isotope-turnover studies. Altered SG plays a major role. Prevention and treatment of early hypoglycemia should be based on adequate glucose replacement. Strategies that reduce insulin secretion or effects appear to be of minor importance.