Using a cost-utility analysis, the effectiveness of tympanostomy tubes was compared to that of antibiotic chemoprophylaxis in young patients with recurrent otitis media. The tympanostomy approach (T-tubes) consisted of placement of a polyethylene grommet in the tympanic membrane, with systemic and local antibiotics administered for one week. The chemoprophylaxis approach consisted of antibiotics in full doses for seven to ten days, followed by continuous antibiotic chemoprophylaxis for six months. Because the T-tube strategy under the model assumptions was more expensive ($396.44 vs $281.30) and yielded slightly less benefit (net utility of .9325 vs. .9476 for initial antibiotic therapy), the chemoprophylaxis option was preferred. We conclude that the initial treatment for recurrent otitis media should consist of acute antibiotics followed by chemoprophylaxis, with T-tubes reserved for treatment failure. Extreme changes in the baseline probabilities of cure or recurrence with antibiotic therapy or in the cost of antibiotic therapy or tympanostomy surgery were required to alter this conclusion. Varying therapy preference (utility) values did not materially alter the conclusions.