We enrolled 60 children with recurrent acute otitis media (AOM) in a study of the effectiveness of antimicrobial prophylaxis. All children were entered into the study following an acute episode of infection treated with amoxicillin (AMX) for 10 days. Following therapy, the children were re-examined, and then randomly assigned to receive either trimethoprim-sulfamethoxazole (TMP-SMX), amoxicillin (AMX) or a placebo (PLA). Twenty children were included in each group. Each drug was administered once a day at bedtime, at 1/3 the therapeutic dose, for 3 months. Children were re-evaluated with pneumootoscopy during episodes of acute illness and with pneumootoscopy and impedance tympanometry (TYMP) at monthly intervals. We observed a significantly increased rate of recurrent AOM in children receiving placebo compared with those who received antibiotics (50% vs. 17% P < 0.005). Both prophylactic antibiotics were equally effective in preventing recurrent AOM (recurrence rate 20% TMP-SMX, 15% AMX). We also observed that recurrences in children receiving placebo occurred earlier in the study period than in those receiving antibiotics. These results suggest that antimicrobial prophylaxis in children with recurrent acute otitis media is effective in reducing subsequent disease. The similar efficacy of both antibiotics tested suggests that the less expensive agent should be used.