The beta-haemolytic streptococcus A, which is responsible for almost all bacterial sore throats in children and young adults, is the target for the first antibiotic treatment. What is really at stake for the treatment is to prevent the serious complications linked with this germ, acute articular rheumatism (RAA) and glomerulonephritis. The recurrence of RAA seen in the United States from 1985, has started many investigations as much for epidemiology as for therapy. Some authors have questioned the treatment with penicillin because of the rise in numbers of carriers of streptococcus A after treatment. They have suggested the use of compounds with a greater spectrum of activity. However, work in the international literature re-states the benefits of a targetted antibio-therapy in this context. Is it appropriate to use a wide spectrum antibiotic to treat an infection of monobacterial target, taking into account the bacteriological and clinical efficacy of penicillin, the only available treatment that has shown its capacity to prevent RAA, with a treatment length of 10 days and which seems to be constant with time? Because of this, faced with the ecological risks of a multi-use of wide range compounds, with a treatment that has not been validated in France, phenoxymethylpenicillin, Oracilline, penicillin V, the antibiotic targetted to sore throat and the streptococcal risk, is correct now for 1994.