Many epidemiologists have agreed that a refined estimate of the incubation period of severe acute respiratory syndrome (SARS) would need a sample size of about 200 cases and appropriate statistical methods enabling the inclusion of cases with defined periods of exposure. However, no such studies have been reported so far. Besides, determinants of the SARS incubation period remain unclear. In this study, 209 probable SARS cases with documented episodes of exposure between March 1 and May 31, 2003, in mainland China were included. A nonparametric method was used to analyze these data with defined periods of exposure to obtain the refined estimate of the SARS incubation period. Furthermore, the authors also explored the influence of various factors on the SARS incubation period by analysis of variance, linear regression analysis, and analysis of covariance. The estimates of mean and variance of the SARS incubation period were 5.29 days and 12.33 days(2), respectively; 90% of patients would have an incubation period of less than 11.58 days with a probability of 0.8, and 99% of patients would have an incubation of less than 22.22 days with a probability of 0.9. The affected area showed a highly significant effect on the incubation period (p < 0.001), but the contact pattern, occupation, gender, and age did not.