Trace element concentrations in plant bioindicators are often determined to assess the quality of the environment. Instrumental methods used for trace element determination require digestion of samples. There are different methods of sample preparation for trace element analysis, and the selection of the best method should be fitted for the purpose of a study. Our hypothesis is that the method of sample preparation is important for interpretation of the results. Here we compare the results of 36 element determinations performed by ICP-MS on ashed and on acid-digested (HNO3, H2O2) samples of two moss species (Hylocomium splendens and Pleurozium schreberi) collected in Alaska and in south-central Poland. We found that dry ashing of the moss samples prior to analysis resulted in considerably lower detection limits of all the elements examined. We also show that this sample preparation technique facilitated the determination of interregional and interspecies differences in the chemistry of trace elements. Compared to the Polish mosses, the Alaskan mosses displayed more positive correlations of the major rock-forming elements with ash content, reflecting those elements' geogenic origin. Of the two moss species, P. schreberi from both Alaska and Poland was also highlighted by a larger number of positive element pair correlations. The cluster analysis suggests that the more uniform element distribution pattern of the Polish mosses primarily reflects regional air pollution sources. Our study has shown that the method of sample preparation is an important factor in statistical interpretation of the results of trace element determinations. ?? 2010 Springer-Verlag.