Fast GC-MS with narrow-bore columns combined with effective sample preparation technique (QuEChERS method) was used for evaluation of various calibration approaches in pesticide residues analysis. In order to compare the performance of analyte protectants (APs) with matrix-matched standards calibration curves of selected pesticides were searched in terms of linearity of responses, repeatability of measurements and reached limit of quantifications utilizing the following calibration standards in the concentration range 1-500 ng mL(-1)(the equivalent sample concentration 1-500 microg kg(-1)): in neat solvent (acetonitrile) with/without addition of APs, matrix-matched standards with/without addition of APs. For APs results are in a good agreement with matrix-matched standards. To evaluate errors of determination of concentration synthetic samples at concentration level of pesticides 50 ng mL(-1) (50 microg kg(-1)) were analyzed and quantified using the above given standards. For less troublesome pesticides very good estimation of concentration was obtained utilizing APs, while for more troublesome pesticides such as methidathion, malathion, phosalone and deltamethrin significant overestimation reaching up to 80% occurred. According to presented results APs can be advantegously used for "easy" pesticides determination. For "difficult" pesticides an alternative calibration approach is required for samples potentially violating MRLs. An example of real sample measurement is shown. In this paper also the use of internal standards (triphenylphosphate (TPP) and heptachlor (HEPT)) for peak areas normalization is discussed in terms of repeatability of measurements and quantitative data obtained. TPP normalization provided slightly better results than the use of absolute peak areas measurements on the contrary to HEPT.