Abstract: Objective To analyze the duplicate reports of pulmonary tuberculosis in Fujian from 2005 to 2009, offering recommendations for improving the report quality. Methods Report cards of the Disease Surveillance Information Report and Management System were analyzed using SAS 9.0. Results Reporting cards with identical patient name, sex, disease, national standard code of current address, where the discrepancy in age was not more than 2 years, were deemed duplicate. The annual constituent of duplicate reports decreased to 0.82% in 2009 from 2.64% in 2005 with an overall declining trend. Duplicate reports were stably present in separate years with a 5-year average incidence of 6.61%. Negatively correlated with the span, the cross-year constituent of duplicate reports averaged 2.39% in a 1-year length and 1.42%, 0.95% and 0.64% in a 2-year, 3-year and 4-year span, respectively. Conclusion Duplicate reports severely compromised the reported incidence rate, thereby entailing a prompt improvement in duplicate finding. Additional duplicate detection indicators and completion of identity card numbers are recommended.