Model fitting and prediction of scarlet fever in the downtown area of Beijing was conducted through time series analysis to describe the epidemiological trend. A database was built and data were fitted with Excel. ARIMA analysis and prediction were made with SPSS. Data from 1957 to 2001 were used for modeling. Data from 2002 to 2004 were used to validate the precision of the model. The incidence of scarlet fever in the downtown area of Beijing since 1957 declined, although fluctuations were apparent. There were two epidemic periods of scarlet fever, at 6.8571 years and 4.8000 years (P<0.10). The incidence in 2008 was predicted as 4.707/100,000 (95% confidence level: 1.379, 16.071; R2 =0.296). Scarlet fever in Beijing is a periodical epidemic. The data of scarlet fever can be analyzed by ARIMA model.