Diabetes mellitus and CVD are some of the leading causes of mortality and morbidity. Accumulating data indicate that a diet characterised by low-glycaemic index (GI) foods may improve the management of diabetes or lipid profiles. The objective of the present meta-analysis was to critically analyse the scientific evidence that low-GI diets have beneficial effects on carbohydrate and lipid metabolism compared with high-GI diets. We searched for randomised controlled trials with a crossover or parallel design published in English between 1981 and 2003, investigating the effect of low-GI v. high-GI diets on markers for carbohydrate and lipid metabolism. Unstandardised differences in mean values were examined using the random effects model. The main outcomes were fructosamine, glycated Hb (HbA1c), HDL-cholesterol, LDL-cholesterol, total cholesterol and triacylglycerol. Literature searches identified sixteen studies that met the strict inclusion criteria. Low-GI diets significantly reduced fructosamine by -0.1 (95 % CI -0.20, 0.00) mmol/l (P=0.05), HbA1c by 0.27 (95 % CI -0.5, -0.03) % (P=0.03), total cholesterol by -0.33 (95 % CI -0.47, -0.18) mmol/l (P<0.0001) and tended to reduce LDL-cholesterol in type 2 diabetic subjects by -0.15 (95 % CI -0.31, -0.00) mmol/l (P=0.06) compared with high-GI diets. No changes were observed in HDL-cholesterol and triacylglycerol concentrations. No substantial heterogeneity was detected, suggesting that the effects of low-GI diets in these studies were uniform. Results of the present meta-analysis support the use of the GI as a scientifically based tool to enable selection of carbohydrate-containing foods to reduce total cholesterol and to improve overall metabolic control of diabetes.