The problems associated with antibiotic resistance have led to several agency and governmental reports since 1998, along with many sets of usage guidelines. These documents stress the desirability of reducing antimicrobial prescribing, which has subsequently fallen in several countries, including the UK. However, the evidence for any contingent reduction in resistance is scanty, and several pathogens--notably Escherichia coli--are becoming markedly more resistant. Thus, rather than being overly optimistic about the benefits of reducing antimicrobial prescriptions, we must also emphasise the use of those antibiotics that prove less prone to select resistance. Furthermore, we must be careful that guidelines are not so narrow as to rail-road prescribing and its contingent selection pressure in single directions--as happened with gonorrhoea--and to consider the likelihood that limited diverse prescribing may have the least detrimental effect upon the resistance ecology. Last, there is a need to re-invigorate antimicrobial development, which has been downgraded by many major pharmaceutical houses.