Insulin resistance is a key element of metabolic syndrome, which includes disturbances of glucose tolerance, obesity, hypertension, coronary heart disease dyslipidemia and many other defects. An important problem in scientific research is precise measurement of insulin sensitivity. The method considered "the gold standard" is glucose clamp, however, it is difficult to apply this method in large studies. Therefore, simple indices of insulin resistance are proposed. It remains unclear whether those indices are able to reflect changes occurring during insulin-sensitizing intervention. The aim of the present study was to assess the use of indirect indices for the changes in insulin sensitivity during exercise training and to compare those indices with results derived from clamp. Fourteen obese normoglycemic women participated in 12-week exercise training program, which included exercise performed on a bicycle ergometer, 5 days a week for 30 minutes. Insulin sensitivity (M/FFM value) before and after training was measured with hyperinsulinemic euglycemic clamp technique. Simple indices of insulin resistance were also assessed: fasting plasma insulin (INS), logarithm INS (log [INS]), homeostasis model assessment (HOMA), logarithm HOMA (log [HOMA]) and quantitative insulin sensitivity check index (QUICKI). Before training, all those indices were markedly related to M/FFM. After training, an increase in M/FFM was observed. None of the examined indices markedly changed after training. There was no correlations between changes of evaluated indices and in M/FFM during training, and no relationships of those parameters after training. Our study indicates that simple indices are not able to reflect changes occurring during insulin-sensitizing intervention.