Abstract: Objective To prove the trend of esophageal cancer mortality from 1970 to 2004 and to provide the evidence for etiology study and preventing strategy of esophageal cancer in Feicheng city.Methods The association among mortality of esophageal cancer and age,gender,periods,birth cohorts were analyzed by negative binomial regression model.Results The results showed that the risk of esophageal cancer mortality associated with age and gender.The risk of esophageal cancer mortality increased by 1167(95%CI:1156～1178)times from above 40 years old group to every other five years old groups.The risk was 2138(95%CI:1178～3118)times higher in male than that in female.There was no significant association between the risk of esophageal cancer mortality with period times or birth cohorts.Conclusion The frequency distribution of the mortality of esophageal cancer was consistent with the negative binomial distribution.The results of analysis by negative binomial regression indicated that the etiology factors for esophageal cancer maybe not change in recent thirty years in the area.