This study assessed the effects of short-term circuit weight training (CWT) on glycaemic control in NIDDM. Twenty-seven untrained, sedentary subjects (mean age, 51) with NIDDM participated in an 8-week randomised, controlled study, involving either CWT 3 days/week (n = 15) or no formal exercise (control) (n = 12). All subjects performed regular self-blood glucose monitoring throughout. Fasting serum glucose and insulin were measured following a 12-h fast and during an oral glucose tolerance test (75 g) before and after 8 weeks. Twenty-one subjects completed the study (CWT, n = 11) (Control, n = 10). Strength for all exercises improved significantly after CWT. Pooled time-series analysis, using a random effects model, revealed an overall decrease in self-monitored glucose levels with CWT compared to controls. Significant reductions from baseline values were observed in both the glucose (-213 mmol l-1 per 120 min, P < 0.05) and insulin (-6130 pmol l-1 per 120 min, P < 0.05) area under the curve following CWT relative to controls. After adjustment for body mass changes, the change in self-monitored glucose levels and insulin area under the curve, but not glucose area under the curve, remained significant. Short-term CWT therefore may provide a practical exercise alternative in the lifestyle management of this condition.