To the prevailing biochemical/physiological classification of mechanisms of organismal resistance to toxicants, an additional molecular dimension is proposed. Predictions are developed regarding the relative prevalence of different classes of mutations and are found to compare favorably with reports from the literature. In particular, point mutations in target loci were the dominant form of resistance for both lab and field selection. Amplifications of target loci were less common than structural mutations, and more common for lab-selected than for field-selected strains. Amplification was the most common mechanism of up-regulation of metabolizing enzymes. In comparison, only one mutation involving cis-regulation and several involving trans-acting regulation were found. Mutations involving gene disruption and down-regulation were uncommon, but were found in appropriate cases, i.e., when toxicants stimulated rather than inhibited target function and when metabolizing enzymes converted toxicants into more toxic metabolites. Additional phenomena of likely but uncertain importance are genetic "succession," recombinational limitation, and negative cross-resistance. More work on these phenomena and on quantification of fitness costs of resistance is recommended.