BACKGROUND Gestational diabetes mellitus (GDM) is a form of diabetes that occurs during pregnancy which can result in significant adverse outcomes for mother and child both in the short and long term. The potential for adverse outcomes, in addition to the increasing prevalence of gestational diabetes worldwide, demonstrates the need to assess strategies, such as dietary advice, that might prevent gestational diabetes. OBJECTIVES To assess the effects of dietary advice in preventing gestational diabetes mellitus. SEARCH STRATEGY We searched the Cochrane Pregnancy and Childbirth Group's Trials Register (January 2008) and reference lists of retrieved articles. SELECTION CRITERIA Quasi-randomised and randomised studies of dietary intervention for preventing glucose intolerance in pregnancy. DATA COLLECTION AND ANALYSIS Two review authors independently conducted data extraction and quality assessment. We resolved disagreements through discussion or through a third author. MAIN RESULTS Three trials (107 women) were included in the review. One trial (25 pregnant women) analysed high-fibre diets with no included outcomes showing statistically significant differences. Two trials (82 pregnant women) assessed low glycaemic index (LGI) versus high glycaemic index diets for pregnant women. Women on the LGI diet had fewer large for gestational age infants (one trial; relative risk (RR) 0.09, 95% confidence interval (CI) 0.01 to 0.69), infants with lower ponderal indexes (two trials; weighted mean difference (WMD) -0.18, 95% CI -0.32 to -0.04, random-effects analysis) and lower maternal fasting glucose levels (two trials; WMD -0.28 mmol/L 95% CI -0.54 to -0.02, random-effects model). Results for women on the LGI diet on neonatal birth weight were not conclusive under a random-effects model (two trials; WMD -527.64 g, 95% CI -1119.20 to 63.92); however, on a fixed-effect model, women on the LGI diet gave birth to lighter babies (two trials; WMD -445.55 g, 95% CI -634.16 to -256.95). High heterogeneity was observed between the trials in most results and both were relatively small trials. One of these trials also included a standard exercise regimen for all participants. AUTHORS' CONCLUSIONS While a low glycaemic index diet was seen to be beneficial for some outcomes for both mother and child, results from the review were inconclusive. Further trials with large sample sizes and longer follow up are required to make more definitive conclusions. No conclusions could be drawn from the high-fibre versus control-diet comparison since the trial involved did not report on many of the outcomes we prespecified.