Acute rhinosinusitis is a common disease, in both children and adult patients, and happens most often in the setting of a viral infection with or without bacterial superinfection. Although spontaneous resolution is common, antibacterials are often prescribed and have a tremendous impact on costs, either directly or through the emergence of resistance in causative or colonising micro-organisms. The purpose of this work was to review published literature from 1989 to 2002 on antibacterial treatment in acute rhinosinusitis from a clinical and economical perspective. A relatively small number of studies have compared antibacterials with placebo and few have suggested that antibacterials are superior to placebo, except when a bacterial cause is established or in the presence of specific CT-scan findings. On the other hand, 58 randomised controlled trials were published between 1989 and 2002, that compared the relative efficacy of various antibacterials. Most of these studies had serious methodological flaws, and no single antibacterial proved superior to its comparators. Economic data are scarce and indicate cost of disease is high. Of the different treatment strategies assessed symptomatic treatment (patients being treated with antibacterials only if they failed to improve after 7 days) was the most cost-effective approach, compared with treating patients on the basis of specific clinical criteria, empirical treatment (all patients initially treated with antibacterials), or radiology-guided treatment. Cost effectiveness varied with disease prevalence. In conclusion, this pharmacoeconomic review of antibacterial use in acute rhinosinusitis shows the need for improvement in the quality of the studies feeding economic analyses, but suggests that huge financial interests are at stake. Savings achievable, by better targeting patients needing antibacterial treatment, could be substantial, and more practical and precise diagnostic procedures are clearly needed. Acute rhinosinusitis is a typical example of a clinical dilemma in which good clinical practice must be balanced against imperfect information and patients' individual interests balanced against society's interest.