Abstract: Objective To analyze the clinical characteristics of pertussis cases and the diagnosis performance in Tianjin during 2010-2015, and provide evidence to improve the clinical identification of pertussis. Methods Descriptive epidemiological analysis was conducted by using the passive surveillance data from hospitals and active surveillance data from communities in Tianjin during this period. Results The median cough duration was 14 days. Cough lasted for less than 14 days in 46.45% of the cases, and 25.63% of the cases showed a persistent cough only without other specific clinical symptoms. The main symptom was paroxysmal spasmodic cough (74.37%), followed by post cough vomiting (32.95%), inspiratory whooping ( 30.00%), cyanosis ( 30.00%) and apnea ( 30.00%). The cases' symptom detection rate in hospital surveillance was significantly higher than that in community surveillance (P 0.05). The proportion of the cases with paroxysmal cough, post cough vomiting, inspiratory whooping, cyanosis and apnea was highest in age group 1 year group, and lowest in age group 15 years. The proportion of the cases with the symptoms mentioned above in the unvaccinated was significantly higher than that in the vaccinated (P 0.05). The misdiagnosis rate of community surveillance was 95.16% in the first medical seeking, in which half of the cases were misdiagnosed with pneumonia. Conclusion The clinical symptoms of pertussis are untypical, and the misdiagnosis rate is high. It is suggested to revise the current case definition of pertussis surveillance to improve the surveillance for pertussis.