The usefulness of the glycaemic index (GI) of a food for practical advice for individuals with diabetes or the general population depends on its reliability, as estimated by intra-class coefficient (ICC), a measure having values between 0 and 1, with values closer to 1 indicating better reliability. We aimed to estimate the ICC of the postprandial blood glucose response to glucose and white bread, instant mashed potato and chickpeas using the incremental area under the curve (iAUC) and the GI of these foods. The iAUC values were determined in twenty healthy individuals on three and four occasions for white bread and glucose, respectively, and for potato and chickpeas on a single occasion. The ICC of the iAUC for white bread and glucose were 0·50 (95 % CI 0·27, 0·73) and 0·49 (95 % CI 0·22, 0·75), respectively. The mean GI of white bread was 81 (95 % CI 74, 90) with a reliability of 0·27 indicating substantial within-person variability. The GI of mashed potato and chickpeas were 87 (95 % CI 76, 101) and 28 (95 % CI 22, 37) respectively with ICC of 0·02 and 0·40.The ICC of the iAUC were moderate and those of the GI fair or poor, indicating the heterogeneous nature of individuals' responses. The unpredictability of individual responses even if they are the result of day-to-day variation places limitations on the clinical usefulness of GI. If the very different GI of potato and chickpeas are estimates of an individual's every-day response to different foods, then the GI of foods may provide an indication of the GI of a long-term diet.