Forty-four children with a clinical diagnosis of streptococcal pharyngitis had throat cultures performed at the initial evaluation and were assigned by randomization to receive either oral penicillin or a placebo for 72 hours. The treating physician, who remained blind to the treatment regimen, recorded the child's temperature and assessed the presence and severity of other signs and symptoms initially and at 24, 48, and 72 hours. The throat culture was positive for group A beta-hemolytic streptococci in 26 (59%) of the initial study group, and most of these children developed a fourfold or greater titer rise in antistreptococcal antibodies in their serum, confirming the diagnosis of streptococcal pharyngitis. Statistically significant clinical improvement was observed in the group of 11 children who were later shown to have been taking penicillin compared with the group of 15 who had taken the placebo. Significant differences in the presence and degree of fever and severity of symptoms persisted in the placebo-treated group for 48 hours. We conclude that early penicillin treatment of children with streptococcal pharyngitis significantly alters the acute clinical course of the disease.