Although the incidence of acute rheumatic fever has declined in the last decades, a few outbreaks have recently been reported. A rapid latex agglutination test for group A streptococci seems reasonably accurate, and early treatment of acute pharyngitis seems to influence the pharyngitis itself. These factors have promoted uncertainty concerning the current best management of patients with sore throat.Clinical decision analysis is used to compare the risks and benefits of symptomatic treatment, and oral and intramuscular penicillin as therapeutic options, and the throat culture and the rapid latex agglutination test as diagnostic strategies. Best estimates of the risk of streptococcal pharyngitis, its complications, the carrier rate, the accuracy of diagnostic tests, the efficacy of antibiotic treatment, allergic reactions, medication compliance, and health outcomes are combined into a management advisory. All results are subjected to a sensitivity analysis in order to check their strength against plausible changes in assumptions. Quality adjusted life days (QALD) lost are used as an outcome measure.The agglutination test combined with oral penicillin yielded the lowest expected loss (.50) of QALD for a typical child with a risk of harboring streptococci of .60. The other strategies, however, yielded losses that were only several hundredths of QALD higher.For children with at least a 40% chance of harboring streptococci and a duration of complaints of less than 2 days before starting treatment, diagnostic testing and prescription of oral penicillin appear to be the best choice of initial management. The rapid latex agglutination test is more effective than the throat culture, because prompt penicillin treatment after a positive test result may shorten the duration of pharyngitis in infected children. High rates of acute rheumatic fever (over 5 X 10(-4] and low medication compliance change the best strategy to agglutination test with intramuscular administration of penicillin.