Background: Advanced glycation end products (AGEs) are associated with diabetes mellitus. Digested food-derived AGEs have been implicated in the pathogenesis of AGE-related disorders, and restricting diet-derived AGEs improves insulin resistance in animal models. The AGE content in foods changes according to cooking method, and it is higher in baked or oven-fried foods than in those prepared by steaming or simmering. Here, we examined the feasibility of crossover comparison tests for determining how different cooking methods (normal diet vs. low-AGE diet) affect insulin levels in non-diabetic Japanese subjects. Methods: Five adult men and women (age, 41 ± 7 years; body mass index (BMI), 21.7 ± 2.6 kg/m2) were enrolled. The following dietary regimen was used: days 1 - 3, control meal; day 4, test meal (normal diet vs. low-AGE diet); day 5, washout day; and day 6, test meal. On days 4 and 6, blood samples were collected before and at 2, 4, and 6 h after meals. Results: Blood levels of N-(carboxymethyl) lysine (CML) increased with dietary intake, but the increase was similar for the normal diet and low-AGE diet groups. Mean plasma glucose, insulin, triglycerides (TG), and CML did not differ significantly between the two groups. The area under the curve (AUC) for insulin levels was lower in the low-AGE diet group (d = 0.8). The sample size calculated from the effect size of the insulin AUC change was 22. Conclusions: Twenty-two subjects may be needed to investigate the changes in clinical parameters attributable to cooking method in non-diabetic Japanese subjects.