Purpose : The accuracy of bacteriologic diagnosis of beta-hemolytic streptococcal pharyngotonsillitis depends on the degree of carrier rate in that area where the throat swabs are obtained and the evaluation of serological T typing as an epidemiologic marker is important to understand epidemiology of streptococcal infection. The purpose of this study is to know the carrier rates of group A streptococcus in normal children form four different areas and to find out the epidemiologic characteristic in distribution of the serotypes for 3 years. Method : Throat swabs were obtained from the tonsillar fossa of normal school children in four different areas(Uljin, Seoul, Osan, Kunsan) from March to May 1996, in Uljin in April 1997, and in Uljin in April 1998. The samples were plated on a 5% sheep blood agar plate and incubated overnight at $37^{\circ}C$ before examination for the presence of beta-hemolytic colonies. All isolated beta-hemolytic streptococcus were grouped and serotyped by T agglutination. Results : The carrier rate of beta-hemolytic streptococci and group A streptococci in 1996 were 27.6%, 18.6% at Uljin; 16.4%, 2.7% at Seoul; 33.0%, 26.0% at Osan; 20.0%, 12.3% at Kunsan, respectively. Among 1,192 normal school children from 4 different areas, we obtained 179 strains of group A streptococci. Fifty two percent of the strains were typable by T agglutination in 1996. Common T-type in 1996 were NT, T1, T3, T2 at Uljin; T12, T25 at Seoul; NT, T6, T28 at Osan; T25, T4, NT, T5 at Kunsan, in decreasing order, respectively. At Uljin, T1, T3, T25 accounted for 69% of strains in 1996, T1, T12, T25 accounted for 70% in 1997, and T12, T4 accounted for 88% in 1998. Conclusion : Higher carrier rates were found in Uljin and Osan, where there are a lower population density with scanty of medical facilities compared with another areas. We supposed that low carrier rates is likely to be related to antibiotic abuse or some epidemiologic factor. The periodic and seasonal serotyping analysis is important in monitoring and understanding the epidemiologic patterns of group A streptococci.