The purpose of this study was to compare the bacteriologic and clinical efficacy of oral penicillin versus amoxicillin as first-line therapy for group A beta-hemolytic streptococcal (GABHS) tonsillopharyngitis. The prospective observational study was conducted over 18 months (January 2000-June 2001). Children enrolled had acute onset of symptoms and signs and a laboratory-documented GABHS tonsillopharyngitis illness. Follow-up examination and laboratory testing occurred 10 +/- 4 days following completion of treatment. In total, 389 patients were enrolled (intent-to-treat group): 195 received penicillin V and 194 received amoxicillin. Fifty-six of the penicillin-treated and 57 amoxicillin-treated patients refused to take the drug, or were noncompliant, or did not return for the follow-up visit, leaving 276 patients in the per-protocol group: 139 penicillin-treated and 137 amoxicillin-treated. Bacteriologic cure for amoxicillin-treated children occurred in 76% versus 64% in the penicillin-treated children (p = 0.04). The clinical cure rate for amoxicillin-treated children was 84% compared to 73% in the penicillin-treated children (p = 0.03). Since treatment allocation was not randomized, logistic regression analysis was used to adjust for treatment group differences. The odds ratio (OR) estimate for cure for patients in the amoxicillin versus penicillin V treatment group remained significant (OR = 1.84, 95% confidence interval 1.02-3.29); the same was true for dinical cure (OR = 1.99, 95% CI = 1.02-3.87). Amoxicillin may be superior to penicillin for bacteriologic and clinical cure of GABHS tonsillopharyngitis.