The optimization of a number ofgas chromatography-mass spectrometry (GC-MS) parameters in order to improve peak area precision through the application of a nonparametric statistical test (the sign test) is demonstrated. As an example, the drug screening method used at the Swedish National Forensic Centre (NFC) is optimized, but in principle, any GC-MS method could be optimized using this approach. The GC-MS parameters investigated were those often overlooked in the optimization process, namely injection volume, liner type, oven temperature program, final oven temperature, MS scan range and MS scan rate. The influence of these parameters on the precision of the peak area responses of 11 different compounds in a test mixture was evaluated using the sign test for pairwise comparison. This nonparametric test provides probability values which facilitate the ranking of parameters according to their influence on peak area variation as well as providing a measure of their statistical significance. This study presents the resulting optimized method and shows that the decreased total variation depended predominantly on liner type and MS scan rate settings. This work also shows that optimization of analytical methods can be achieved using simple and easily accessible statistical tools.