The aim of diabetes management is to normalise blood glucose levels since improved blood glucose control is associated with fewer complications. Food affects blood glucose levels; however, there is no universal approach to the optimal diabetic diet and there is controversy about the usefulness of the low-glycaemic index (GI) diet. To assess the effects of low-GI diets on glycaemic control in diabetes, we conducted electronic searches of the Cochrane Library, MEDLINE, EMBASE and CINAHL. We assessed randomised controlled trials (RCT) with interventions >4 weeks that compared a low-GI diet with a higher-GI diet for type 1 or type 2 diabetes. Twelve RCT (n 612) were identified. There was a significant decrease in glycated Hb (HbA1c) with low-GI diet than with the control diet, indicating improved glycaemic control (seven trials, n 457, weighted mean difference (WMD) − 0·4 % HbA1c, 95 % CI − 0·7, − 0·20, P = 0·001). In four studies reporting the results for glycaemic control as fructosamine, three of which were 6 weeks or less in duration, pooled data showed a decrease in fructosamine (WMD − 0·23 mmol/l, 95 % CI − 0·47, 0·00, P = 0·05), n 141, with low-GI diet than with high-GI diet. Glycosylated albumin levels decreased significantly with low-GI diet, but not with high-GI diet, in one study that reported this outcome. Lowering the GI of the diet may contribute to improved glycaemic control in diabetes.