Current international guidelines for the management of community acquired pneumonia (CAP) recommend therapy with a beta-lactam plus a macrolide or a 'respiratory' fluoroquinolone alone in patients hospitalized in a medical ward, and combination therapy with a beta-lactam plus a macrolide or a fluoroquinolone in patients hospitalized in the intensive care unit. However, which of the available options should be preferred remains a matter of debate, and there are surprisingly few prospective randomized trials strictly comparing mono- versus dual therapy strategies in CAP patients. Thus, the recommendation of combining a macrolide with a beta-lactam rather than using a beta-lactam alone in hospitalized patients is derived mainly from observational data, and the suggested combination of a beta-lactam with a fluoroquinolone in severe CAP has been rarely examined in a clinical trial.As there have been sound theoretical arguments for and against combination therapy regimens, the rationale for the different options is discussed and available clinical trial data are reviewed in this article. A final conclusion about the superiority of one antibacterial regimen over another in hospitalized patients with CAP cannot be drawn on the basis of the limited data available. So far, combination therapy probably should be preferred in all patients presenting with severe pneumonia, whereas in general, combination therapy is not necessary in patients in a medical ward, and combination therapy with a beta-lactam plus a macrolide or monotherapy with a respiratory fluoroquinolone should be considered equivalent in this latter patient group. On the other hand, the available data demonstrate that empirical coverage of atypical bacteria in all patients with mild-to-moderate CAP seems unnecessary, and beta-lactam monotherapy might perform equally well when compared with respiratory fluoroquinolones in patients with non-severe CAP. Thus, the alternative use of a beta-lactam alone at adequate dosage in clinically stable patients seems justified, if CAP due to Legionella pneumophila is unlikely.