The significant progress achieved in modern instrumental analysis has led to a continuous lowering of detection limits and improved precision. This should in principle permit the reliable and extremely precise analysis of trace compounds mainly trace elements, at levels down to the lowest natural concentrations. However, the frequently observed very high discrepancies between the analytΒ ical results of different laboratories as well as the deviations from true values are, regrettably, still common in analytical practice. Basic methodological errors at the determination step can usually be minimized or even avoided by carefully perΒ formed quality control measures - e. g. by interlaboratory comparisons and the proper use of certified reference materials. The most severe and often underΒ estimated error sources, however, are those connected with the whole and often extremely complex sampling process, and also to a lesser extent, with sample preΒ paration prior to analysis. Thus, for these initial steps of an analytical procedure particular experience is needed, as well as a detailed knowledge of the interrelaΒ tions between these steps, which always have to be applied with the utmost care. In collaboration with a number of very experienced colleagues working in difΒ book has tried to contribute to a better ferent research areas, the editor of this understanding of these particular error sources and how they can be overcome in a series of training courses held during the last decade at the "Haus der Technik", Essen, Germany.