OBJECTIVE--To determine whether cephalexin or penicillin is more effective in the treatment of group A beta-hemolytic streptococcal tonsillopharyngitis in children. DESIGN--Randomized, double-blind, crossover study conducted from 1981 to 1984. SETTING--Seven pediatric practices in the United States, including private offices and pediatric clinics. PARTICIPANTS--Of the 654 patients, 525 children and adolescents with clinical evidence of tonsillitis or pharyngitis and throat cultures positive for group A beta-hemolytic streptococcal infection were evaluable. Eighty percent of patients completed the study; none were withdrawn because of adverse reaction. SELECTION CRITERIA--Children and adolescents who had acute illness suggestive of group A beta-hemolytic streptococcal infection were enrolled in the study. Treatment was continued if the throat culture was positive for group A beta-hemolytic streptococcal infection. INTERVENTIONS--Four doses of cephalexin and penicillin (27 mg/kg per day) were prescribed to be taken on an empty stomach for 10 days. MEASUREMENTS/MAIN RESULTS--Symptomatic clinical failure occurred in 8% of penicillin-treated patients and in 3% of cephalexin-treated patients. Bacteriologic failure rates were 11% in the penicillin treatment group and 7% in the cephalexin treatment group. The combined treatment failure rate of clinical relapse plus asymptomatic bacteriologic failure was 19% in the penicillin treatment group and 10% in the cephalexin treatment group. Paired antistreptolysin-O titer increased significantly in 62.3% of penicillin-treated patients and in 64.2% of cephalexin-treated patients. Similarly, anti-DNase B titers rose 52.2% in penicillin-treated patients and 52.4% in cephalexin-treated patients. CONCLUSION--Cephalexin is a more effective drug than penicillin in the treatment of group A beta-hemolytic streptococcal throat infection in children.