Background.Three-day, 10 mg/kg/day azithromycin (AZM) studies in pediatric acute group A streptococcal tonsillopharyngitis have shown contradictory bacteriologic results. This study investigates the efficacy and tolerability of two dosages of 3-day azithromycin (20 mg/kg/day and 10 mg/kg/day) compared with 10-day penicillin V.Methods.This was a prospective, comparative, randomized, multicenter trial. Children were scheduled to return for visits at 14 days (main end point) and 1 month after the onset of treatment for clinical and bacteriologic assessment. Molecular tools were used to compare pre- and posttreatment group A beta-hemolyticStreptococcus(GABHS) isolates.Results.Between November, 1997, and July, 1998, 501 patients (169 AZM 10 mg, 165 AZM 20 mg, 167 penicillin V) between 2 and 12 years old were enrolled; 500 were assessable for safety, 469 for intent to treat analysis and 420 for efficacy in the per protocol analysis. Before treatment 25 (7.9%) of 315 GABHS stains isolated from patients receiving AZM were resistant to this compound. On Day 14 pretreatment GABHS were eradicated from 78 (57.8%) of the 135 children receiving the AZM 10 mg regimen, 131 (94.2%) of the 139 receiving AZM 20 mg and 123 (84.2%) of the 146 taking penicillin. One month after the outset of treatment, bacteriologic relapses were observed in 40.5% (n= 30) of the children receiving AZM 10 mg, 14.8% (n= 18) of children taking AZM 20 mg and 13.2% (n= 15) of those treated with penicillin V. AZM 20 mg/kg/day was statistically superior to AZM 10 mg/kg/day microbiologically on Day 14 (P= 0.0001) and Day 30 (P= 0.0001) and clinically on Day 14 (P= 0.0035). AZM 20 mg/kg/day was statistically equivalent both microbiologically and clinically to standard therapy with penicillin V at all endpoints. The incidence of treatment-related adverse events was similar in the two azithromycin groups [AZM 10 mg, 31 of 169 (18.3%); AZM 20 mg, 37 of 164 (23%)] but significantly higher than those observed in the penicillin V group [5 of 166 (3%);P< 0.0001]. Most treatment-related adverse events were gastrointestinal and of mild-to-moderate severity. Fourteen patients withdrew from the trial because of adverse events (1 in the penicillin V group, 7 in the AZM 10 mg group and 6 in the AZM 20 mg group).Conclusion.This is the first study to demonstrate a daily dose-dependent difference in microbiologic efficacy of a regimen; 3-day AZM 20 mg/kg/day is a more effective regimen than 3-day AZM 10 mg/kg/day for pediatric GABHS tonsillopharyngitis.