Despite the fact that β-lactamases from a range of bacterial species - Gram-positive and Gram-negative - show evolutionary relatedness, there is no set pattern to the genetic organization that underlies their synthesis and its regulation. Thus, for example, the enzymes of many Gram-positive species are extracellular and inducible, whereas their counterparts in Gram-negative bacteria are often produced constitutively into the periplasmic space of the cells concerned. Nor is the location of the β-lactamase genes always the same: in Escherichia coli and Staphylococcus aureus , for example, these are commonly plasmid-borne, whereas with other species the genes are chromosomal. Furthermore, the location may not be fixed: some strains of a species may, for example, have their β-lactamase genes on a plasmid, whereas others of the same species may carry the same genes as part of their chromosome. In many cases the highly flexible genetic arrangement that underlies β-lactamase synthesis derives from two main features: first, where plasmids are involved, their ability to be transferred to related species, and the fact that they can often replicate in their new hosts, ensure that the genes specifying a given type of β-lactamase may move from species to species. Thus one finds enzymes of the same type in many distinct strains and species. The second source of flexibility is that the gene concerned is often part of a transposon: a genetic element incapable of independent replication, but which can move from one bacterial replicon to another by a mechanism independent of normal generalized recombination. Thus, with many β-lactamases - as also with enzymes that inactivate other antibiotics - their genes may move from replicon to replicon within a given bacterial cell, and from cell to cell within a bacterial population. This, then, is an arrangement of much evolutionary potential: something which is operated upon by selection pressure to give rise to the resistant bacterial populations which cause so much trouble in our hospitals. In this context, moreover, one can even think of a third level of organization where plasmid-carrying bacteria move from one host to another by a process of cross-infection. Even though it is clear that some β-lactamase genes can spread rapidly in susceptible bacterial populations, there also exist mechanisms that limit the extent to which the spread of both plasmids and transposons occurs. For example, some strains are poor recipients for certain types of plasmid, and, at a lower level of organization, some plasmids are relatively immune to the transposition of β-lactamase transposons. Overall, therefore, as common with evolutionary systems, the performance of the system as it affects β-lactam resistance is dependent on a balance of positive and negative influences: transfer of plasmids and transposons, on the one hand, and immunity to such transfers on the other.