OBJECTIVES: The CAP-IN study compared two methods of treatment for moderately to severely ill patients hospitalized with community-acquired pneumonia: a 7-10 day course of sequential therapy with intravenous azithromycin plus intravenous ceftriaxone followed by oral levofloxacin (AZITH) versus a 7-14 day course of intravenous levofloxican followed by oral levofloxacin (LEVO). The trial was stopped early, with n = 110 for AZITH and n = 102 for LEVO, when similar clinical efficacy was demonstrated. We compared the 30-day direct medical costs of the two treatments. METHODS: The primary inputs to the economic analysis were drug dosage, hospital days and location during the stay, and post-discharge utilization (home care, outpatient visit, emergency department, physician office). For each, observed utilization was multiplied by an estimated unit price. The sensitivity of the results to the variability of the patients in the trial was assessed using bootstrapping. RESULTS: The difference in total cost was $2034, with 95% confidence interval $263 to $3803 (p = 0.025), in favor of AZITH. This result was primarily due to an approximately 1.5 day difference in length of hospital stay. CONCLUSIONS: Direct medical costs for AZITH were $2034 less than the corresponding costs for LEVO, the primary reason being a reduction in the length of stay. AZITH is no more costly than LEVO, and perhaps less so. In selecting between these drugs, physicians should consider various factors, among them being convenience of administration and cost.