aim and methods: our main aim was to determine the association between clinical, demographical parameters and different insulin resistance and secretion indices in apparently healthy subjects, without previous knowledge of their own level of glucose tolerance. for that purpose, we evaluated 105 individuals from february to august 2003 by means of ogtt, aged 33.4 ± 1.4 years old, 57.1% female. we allocated them in four groups: group 0 (normal): individuals with bmi < 25 kg/m2 and normal glucose metabolism, group 1 (obese): bmi > 25 kg/m2 and normal glucose metabolism, group 2 (ifg): impaired fasting glucose and group 3 (igt): impaired glucose tolerance. results: we have found statistical difference on all variables during ogtt between all groups: fasting glucose (p < 0.05), 2-hour glucose (p < 0.05), glucose peak value (p < 0.05), glucose delta (p = 0.02), glucose incremental percentage (p = 0.047), area under curve (p < 0.05), and glucose peak time (p = 0.022). we have not found difference on any variable in insulin curves or on glucose incremental velocity. regarding insulin secretion indices there were no statistical significance in insulinogenic or delta indices, but they became significant after being corrected by insulin resistance (p = 0.008). when we evaluated insulin resistance alone, by using homa and quicki indices and the fasting glucose to insulin index, we have found statistical significance (p = 0.005; p = 0.005; p = 0.053). conclusion: although studying a small sample, we could suggest that individuals with impaired fasting glucose and impaired glucose tolerance are in different stages of diabetes natural history disease. we found out that the best indices of insulin resistance are both homa and quicki. we also suggest that pancreatic secretion indices should be corrected by the insulin resistance, which could best reflect type 2 diabetes natural history.