Drying oils identification in art objects is an important step in the scientific investigation of the artifact which provides conservators and art historians with valuable information concerning materials used and painting techniques applied. The present communication is devoted to pitfalls and troubleshooting in drying oils identification by means of GC-MS analysis of fatty acids composition in a microsample of an art object. We demonstrate that in the case of nonlinear instrument response the ratios of palmitic to stearic (P/S), distinctive for each oil type and used for drying oil identification, depend on sample dilution so that different dilutions of the same sample can give different P/S ratios. This phenomenon can hinder drying oil identification and lead to erroneous interpretations. This is an important observation as nowadays very often the P/S ratio is calculated from the corresponding peak area ratios or by the use of one-point calibration method. In these approaches, the linearity of the instrument response is not controlled and ensured. In the case analyzed, the nonlinear instrument response was attributed to incomplete sample evaporation in the injector. Packing of the glass liner with deactivated glass wool improved the sample evaporation and ensured the linearity of the instrument response and independence of the P/S ratio from sample dilution.