The effectiveness of continuous compared with intermittent amoxicillin prophylaxis administered to subjects with a history of recurrent otitis media enrolled during the winter respiratory infection season was assessed in a prospective single blinded clinical trial. Patients with three or more chart-documented episodes of otitis media (OM) in the preceding 6 months were enrolled and randomly assigned to each treatment group. Patients in the continuous group received amoxicillin twice a day every day for up to 4 months. During the same period patients in the intermittent group received amoxicillin twice a day only when they developed respiratory symptoms of congestion, runny nose or cough. Among the 30 patients receiving continuous amoxicillin who were followed for at least 3 months, 22 (73%) had no OM episodes and 8 (28%) had one OM episode. Among the 25 patients receiving intermittent amoxicillin for at least 3 months, 13 (52%) had no OM episodes, 8 (32%) had 1 episode and 4 (16%) had 2 episodes. Significantly fewer patients had fewer than 2 OM episodes on continuous compared with intermittent amoxicillin (P less than 0.04). The incidence density was 0.46 episode/120 days at risk in the continuous treatment group compared with 1.10 episodes/120 days at risk for intermittent treatment (P less than 0.03). Among patients 12 months or older the incidence density of OM episodes per 120 days was 3.5 times higher in the intermittent amoxicillin group (0.80) compared with the continuous amoxicillin group (0.23) (P = 0.05). The incidence densities of the continuous vs. intermittent therapy groups did not differ significantly for patients younger than 12 months of age. The findings suggest that continuous amoxicillin prophylaxis may be more effective than intermittent treatment in preventing OM episodes in patients 12 months or older with a history of recurrent otitis media.