OBJECTIVE: To build a utility-based model comparing physicians' prescribing practices for acute sinusitis, assuming that utility is a function of repeat visits, second prescriptions and call-backs. METHODS: An algorithm is applied to data from an open-label, multicenter, randomized trial, which documents call back times, office visits and second prescriptions for 792 Acute Bacterial Sinusitis patients treated with Amoxicillin clavulanate 875mg, Levofloxacin 500mg or Moxifloxacin 400mg. The marginal effect of fewer call-backs, office visits and second prescriptions, using utility as a function of these, is represented by a decision tree. Probabilities from the trial are used. Utilities range from highest (1.00) when the patient does not call back or get a second prescription, to lowest (0.25), when the patient calls back/comes back for a visit. Intermediate utilities range in decreasing order from call-back/no action, to no call-back/second prescription, to call-back for other reasons, to call-back/second prescription. Regardless of the drug prescribed, the utility to the physician of a given outcome is the same. RESULTS: Fewer call-backs and repeat visits pursuant to moxifloxacin resulted in the highest expected value (0.90), compared to levofloxacin (0.89) and Amoxicillin (0.87), when the lowest utility was 0.25. A sensitivity analysis showed consistent results, with 0.86 for Moxifloxacin, 0.85 for Levofloxacin and 0.82 for Amoxicillin when the lowest utility was zero. When the lowest utility was 0.75, all drugs yielded 0.96. CONCLUSION: Physicians who seek to reduce events of call-backs, repeat visits and second prescriptions may favor certain antibiotic regimens, such as Moxifloxacin, for the treatment of acute sinusitis. This is more relevant when physicians put a higher premium on reduced events, that is when the spread in utilities is larger (0.25 and 1.00), than when the spread is smaller (0.75 and 1.00).