Objective To compare results of healthcare -associated infection(HAI) prevalence rate survey conducted through bedside survey and medical records consulting (referred to as manual survey) and HAI real-time monitoring system (referred to as RT-HAI survey), so as to provide reference for improving the investigation method and quality of HAI. Methods From 2014 to 2016, according to cross-sectional survey theory, prevalence rate of HAI in a hospital was investigated by manual survey and RT-HAI survey methods respectively, results were analyzed statistically. Results From 2014 to 2016, a total of 4 821 inpatients were investigated, the prevalence of HAI by manual survey were 3.98%, 3.94% and 3.61% respectively, prevalence of HAI by RT-HAI survey were 4.35%, 4.63% and 4.14% respectively, there was no significant difference in the prevalence of HAI between two methods in the same year (all P>0.05). After analyzed and discussed by HAI control expert group, the actual number of HAI on the survey day from 2014 to 2016 were 70, 68 and 75 cases respectively, with the prevalence rates of 4.22%, 4.70% and 4.37% respectively. Manual survey result and RT-HAI survey result were highly consistent with actual HAI on the same day of the annual survey, with statistical significance (all P<0.05). The sensitivity, Youden index, total coincidence rate and odd product of RT-HAI survey were all higher than those of manual survey, missed diagnosis rate was lower than that of manual survey. Conclusion RT-HAI is advantageous to improve HAI surveillance level.