In their review (low-glycaemic index diets and body weight regulation (2006)), McMillan-Price and Brand-Miller argue that the low glycemic index (GI) diet is a simple and more popular diet that will successfully improve cardiovascular risk factors and reduce body weight. We do not find that there is convincing evidence in the existing literature to suggest that a low GI diet is superior in achieving improvement in cardiovascular health and in reducing body weight in healthy overweight subjects, when compared to official dietary advice recommending a diet high in vegetables, fruit and fiber, and low in sugar and fat. This lack of evidence might partly be due to the lack of long-term, well-powered studies, with well-controlled diets differing only in GI. Data also suggest that subjects' insulin sensitivity might be an important predictor of the effects of a low GI diet, and therefore findings from studies in insulin-resistant and diabetic subjects should not be extrapolated to findings in healthy individuals. We agree with McMillan-Price and Brand-Miller when they state that 'in practice it is difficult to tease out the separate effects of GI, palatability, volume, fiber and other factors that influence satiety responses to realistic meals'. Predicting GI of realistic meals has also proven difficult. We therefore find that future studies should focus on individual food factors, such as the effect of whole grain, including intact grains, fiber, including resistant starch, energy density and preparation methods. This approach would allow for more tightly controlled trials, with less confounding factors, and also lead to simpler dietary advice with assured efficacy.