The effects of regularly eating sucrose were studied in 23 diabetic patients, 12 Type 1 (insulin-dependent) and 11 Type 2 (non-insulin-dependent), with differing degrees of glycaemic control. Two diets, each lasting 6 weeks, were compared in a randomised cross-over study. Both diets were high in fibre and low in fat. In one diet 45 g of complex carbohydrate was replaced by 45 g of sucrose taken at mealtimes. There were no significant biochemical differences between the two diets in either Type 1 or Type 2 patients. In Type 1 patients the mean (+/- SEM) fasting plasma glucose was 10.5 (1.8)mmol/l on the control diet and 10.3 (1.5) mmol/l on sucrose. In Type 2 patients the levels were 9.1 (0.8) mmol/l and 8.9 (0.8) mmol/l respectively. Glycosylated haemoglobin for the Type 1 patients was 9.9% on control and 10.3% on sucrose; for Type 2 patients the figures were 9.3% and 9.0% respectively. There were no differences in mean daily plasma glucose levels or diurnal glucose profiles. Cholesterol (total and in lipoprotein fractions) was unchanged, as were diurnal triglyceride profiles and plasma insulin profiles in the Type 2 patients. There were no changes in medication or body weight. We conclude that a moderate amount of sucrose taken daily at mealtimes does not cause deterioration in metabolic control in diabetic patients following a high fibre/low fat diet.