Streptococus pneumoniae represents an interesting model to discuss the relative impact of broad versus narrow spectrum antibacterials as potential selectors for resistance. Indeed, this pathogen is responsible for potentially severe infections in the community, and has a great capacity for acquisition of resistance to antibacterial agents. It has been the focus of many studies to elucidate some unique aspects of molecular biology, including the adaptive mechanisms responsible for emergence and spread of multiresistance. In the past, the use of narrow spectrum agents was recommended in order to try to reduce the risk of selection of resistance. This concept is nowadays somewhat obsolete for several reasons. S. pneumoniae is able to acquire resistance to antibacterials belonging to different families of drugs through different molecular mechanisms. Thus, selection of multiresistant pneumococci can result from exposure to very different agents, including narrow spectrum as well as broad spectrum agents. In vitro studies have shown a different potential for selection of resistance among the beta-lactam agents. Furthermore, several studies have more or less directly established a close relationship between the level of antibacterial use and the rate of selection of resistance. In addition to the overall amount of antibacterials prescribed in the community, several other factors have been shown to influence the rate of selection of resistance, including the use of doses that are too low, the length of therapy and the duration of bacterial exposure to long-acting agents compared to drugs with short half-lives. Therefore, there are three main ways to control selection and spread of resistant strains: by (i) reducing the amount of antibacterials used; (ii) using optimal dosages (avoiding underdosing) and treatments of short duration; and (iii) reducing the risk of transmission among young children attending daycare centres or kindergartens. In order to help physicians reduce the number of unnecessary prescriptions, it is important to develop rapid tests to recognise the bacterial origin of a febrile illness and even more important to detect resistance to antibacterials. However, apart from rapid diagnostic tests for streptococcal pharyngitis, those tests are not currently available. As a consequence, currently, the debate around narrow versus broad spectrum antibacterials remains a false debate. Physicians should use broad spectrum agents in many instances of upper or lower respiratory tract infection, taking into consideration the probable pathogens and the risk of (multi)resistance to antibacterials. Once rapid diagnostic are available in community practice, allowing a precise diagnosis of the offending agent and its susceptibility profile, physicians will be able to add to their current criteria the selective potential for resistance of the antibacterials that appear to be active in vitro.