In the field of elemental analysis inductively coupled plasma mass spectrometers (ICP-MS) have the best sensitivity that means the lowest limit of detection, subsequently their applicability for the detection of essential and toxic elements in foods and foodstuffs is prominent. For the most elements could be measured the detection limit is between μg kg^-1 (ppb) and ng kg^-1 (ppt) e.g. for arsenic and selenium.

Considering an analytical task (sample type, analytes and their concentration, pretreatment procedure etc.) the applicability of an analytical method is determined by its performance characteristics. The purpose of validation is to ensure that the method would be used fulfills the requirements of the given task. In this article we describes one of the performance characteristics, the linearity, and the whole validation procedure aims measurement of arsenic and selenium in foodstuffs by inductively coupled plasma mass spectrometer (Thermo XSeries I.); but because of the limited number of pages the results are demonstrated only for arsenic.

The linearity of calibration was evaluated in three concentration ranges (0.1–1 μg l^-1; 1–10 μg l^-1; 10–50 μg l^-1), with nine line-fit possibilities (without weighting, weighting with absolute or relative deviation; with or without forcing the curve through blank or origin) and different methods (graphical examination, correlation coefficient, analysis of variance).

The best method to ensure the linearity of correspondence between signal and concentration was the ANOVA test. In view of calibrations it was found that the range of 10–50 μg l-1 could be regarded as linear with four line-fit possibilities, and was non-linear between 0.1–1 μg l^-1 and 1–10 μg l^-1.