Objective To probabilistically estimate the epidemiological burden of nontyphoidal Salmonella infection in Shanghai and to provide decision-making basis for enacting appropriate foodborne diseases prevention policy. Methods Sixty streets/towns in Shanghai were randomly sampled, and annual incidence and hospital-visiting rate of people with acute gastroenteritis were obtained by questionnaire survey. Multi-stage stratified random sampling was used to extract 23 hospitals, and patient stool sampling rate, laboratory testing rate, and positive rate of Salmonella were obtain. Positive case reporting rate of Salmonella was obtained from National Notifiable Disease Reporting System (NNDRS). Annual incidence and epidemiological burden of nontyphoidal Salmonella were calculated by Monte-Carlo simulation according to the epidemiological burden pyramid model, and the active surveillance and passive report system were compared. Results The annual incidence of acute gastroenteritis among residents in Shanghai was 23 261 cases per 100 000, and hospital-visiting rate was 20.47%(26/127). Patient stool sampling rate from surveillance hospital was 21.66%(5 974/27 587),and positive rate of Salmonella was 3.53%(211/5 974). Positive cases reporting rate of Salmonella in active surveillance was about 81.04%(171/211).The annual incidence of passive reporting system was estimated only 5 cases per 100 000. It was probabilistically estimated that the annual incidence of nontyphoidal Salmonella from passive report system was 270 cases(95%CI:185-399)per 100 000, with 0.062 1 million cases(95%CI:0.042 5-0.091 8)and the total underestimation coefficient was 60∶1. While the estimated annual incidence of nontyphoidal Salmonella from active surveillance was 686 cases(95%CI:549-843)per 100 000, with 0.157 8 million cases(95%CI:0.126 4-0.194 2)and the total underestimation coefficient was 154∶ 1. After comparison, the estimates from active surveillance were 2.54 times to passive report system. The incidence of Salmonella infection had obvious seasonal character. The speak season was from May to August. Conclusion The estimated annual incidence of nontyphoidal Salmonella infection from foodborne diseases active surveillance was much higher than that from passive report system. It showed that active surveillance had a good advantage for better understanding of the morbidity, hospital-visiting and epidemiological burden estimation for foodborne diseases. At the same time, the hospital-visiting rate was the biggest factor affecting the total cases of Salmonella infection. This study further demonstrated that epidemiological burden pyramid model was feasible for foodborne diseases in china, especially single category diseases.