The development of resistance to antibiotics by reductions in the affinities of their enzymatic targets occurs most rapidly for antibiotics that inactivate a single target and that are not analogs of substrate. In these cases of resistance (for example, resistance to rifampicin), numerous single amino acid substitutions may provide large decreases in the affinity of the target for the antibiotic, leading to clinically significant levels of resistance. Resistance due to target alterations should occur much more slowly for those antibiotics (penicillin, for example) that inactivate multiple targets irreversibly by acting as close analogs of substrate. Resistance to penicillin because of target changes has emerged, by unexpected mechanisms, only in a limited number of species. However, inactivating enzymes commonly provide resistance to antibiotics that, like penicillin, are derived from natural products, although such enzymes have not been found for synthetic antibiotics. Thus, the ideal antibiotic would be produced by rational design, rather than by the modification of a natural product.