The effect of dietary education incorporating information about the glycaemic index of carbohydrate was tested against standard dietary advice in a randomized controlled study in 51 newly diagnosed patients with Type 2 diabetes treated as out-patients with diet only over a 12-week study period. Outcome was assessed by dietary analysis of 3-day diet diaries, fasting blood glucose, fructosamine, total cholesterol, LDL-cholesterol, HDL-cholesterol, and triglycerides. Dietary analysis indicated that the group who received low glycaemic advice not only had a significantly lower calculated mean diet glycaemic index intake (77 +/- 1.1 (SEM) vs 82 +/- 1%, p < 0.01) but also had a lower fat intake (25 +/- 1 vs 32 +/- 2% of total energy day-1, p < 0.001), a higher carbohydrate intake (49 +/- 2% vs 44 +/- 1% of total energy day-1, p < 0.05) and non-starch polysaccharide intake (21 +/- 1.5 vs 14 +/- 1 g, p < 0.01). There was a significantly greater within-group fall in fructosamine (3.8 +/- 0.2 to 3.2 +/- 0.2 mmol-1 vs 3.6 +/- 0.2 to 3.6 +/- 0.3 mmol-1, p < 0.05) and cholesterol (6.1 +/- 0.3 to 5.4 +/- 0.3 mmol-1 vs. 5.6 +/- 0.2 to 5.3 +/- 0.1 mmol-1, p < 0.05) in the low glycaemic index group.(ABSTRACT TRUNCATED AT 250 WORDS)