The results of a cohort study of 92,346 labour population for probing the relationship between levamisole (LMS) and encephalitis syndrome was reported. There were five patients suffering from the disease in an exposed group (E) of 10,911 people within two months after taking LMS as an anthelmintic for mass treatment, the incidence of the disease is 4.58/10,000. The nonexposed population was divided into two control groups the first group (N1) includes 37,990 people living in the mass treatment villages and the other group (N2) consists of 43,445 people living in the adjoining villages without mass treatment. None in the two control groups suffered from the syndrome or any other encephalitis-like disease within the same period. Statistical tests for the differences between E and either N1 or N2 separately is highly significant. (between E and N1: P = 0.000553, P < 0.001; between E and N2: P = 0.000325, P < 0.001). The overall attributable risk (AR) is 45.8/100,000 and that for female is 93.8/100,000. However, the difference between the incidence rates of the disease in female and male is not statistically significant (P > 0.025). It is shown that LMS is obviously the causative factor of the disease.