1. Modelling analysis of intravenous glucose tolerance test glucose and insulin concentrations can provide measures of insulin sensitivity and metabolism from a single straightforward procedure. However, little is known of the effects of blood arterialization on model-derived parameters. 2. Intravenous glucose tolerance tests were carried out on 18 subjects, with measurement of glucose and insulin concentrations in simultaneously sampled non-arterialized and arterialized blood. Blood oxygen saturation, partial pressure of CO2 and pH were measured on both non-arterialized and arterialized blood during the intravenous glucose tolerance test. Using the minimal models of glucose disappearance and post-hepatic insulin delivery, measures of insulin sensitivity, glucose-dependent glucose disposal, first- and second-phase post-hepatic insulin responsiveness to glucose and plasma insulin elimination rate were derived from intravenous glucose tolerance test glucose and insulin concentrations in both arterialized and non-arterialized blood. 3. During the intravenous glucose tolerance test mean blood oxygen saturation was 6.7% higher, partial pressure of CO2 was 0.3 kPa lower and pH was 0.015 higher in arterialized than non-arterialized blood. Mean parameter values did not differ when derived from measurements made on non-arterialized and arterialized blood. Model-derived parameters were not related to the degree of arterialization, although there was some consistent variation with sampling site for parameters of glucose-dependent glucose disposal (Sg), first-phase post-hepatic insulin responsiveness (phi 1) and insulin elimination (ni). 4. Measurements made on non-arterialized blood are suitable for analyses employing the minimal models of glucose disappearance and post-hepatic insulin delivery. Imprecision in some parameters may be diminished by adherence to a single sampling site.