Phytochemists continue to be amazed (and annoyed) by the fact that one and the same taxon may harbour several biogenetic groups of micromolecules, and they have consequently reached several rather discouraging opinions. According to the most pessimistic one the distribution of micro- molecules is too haphazard to make them systematically relevant. Others, somewhat more optimistically, contend that, while the consideration of only one biogenetic group may be of doubtful value, only the integration of all chemical characters into the system will lead to useful results. Both opinions are wrong. The very variability of natural products, within certain genetic constraints, is the reason for their existence and the explanation for their role in plant life. This point is taken up explicitly in other chapters through the introduction of the geographical parameter into my reasoning. Integration of all data is another matter. The notion that a classification will become more natural as the number of characteristics considered increases certainly overlooks the fact that it would be difficult to consider together compounds with different or unknown functions. However when for instance allelochemics are selected for classificatory purposes, all biogenetic groups with similar functions must be integrated in order to obtain a complete evolutionary profile of the taxon.