In subjects with impaired glucose tolerance hyperproinsulinaemia has been shown to be predictive for progression to Type II (non-insulin-dependent) diabetes mellitus. These findings are often interpreted as early indicators of an impaired beta-cell function. The aim of our study was to assess the potential determinants of hyperproinsulinaemia in subjects with impaired glucose tolerance. The study group consisted of 110 subjects, 45-74 years of age with mean 2 h plasma glucose concentrations between 8.6 and 11.1 mmol/l following two oral glucose tolerance tests. Subsequently, the hyperglycaemic clamp technique (10 mmol/l, with a priming infusion of 20% glucose solution, 150 mg/kg) was used to assess the beta-cell function (time needed to reach the insulin peak) and insulin sensitivity (M/I value: glucose metabolised divided by insulin response, 150-180 min). Results showed that the intact-proinsulin:insulin ratio increased with increasing time needed to reach the insulin peak (0.065, 0.079 and 0.101; time needed to reach the insulin peak < or = 5 min, 5 to 15 min, > 15 min; p < 0.05). The split-proinsulin:insulin ratio showed a similar association with the time needed to reach the insulin peak. These associations were independent of age, sex, body mass index and waist:hip ratio. In conclusion, this study shows that relative hyperproinsulinaemia is associated with an impaired beta-cell function in a study group of subjects with impaired glucose tolerance selected after two oral glucose tolerance tests.