Background The seroprevalence of immunoglobulin M (IgM) and immunoglobulin G (IgG) antibodies against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) may be a more reliable approach to detect true infected population, particularly in asymptomatic persons. Few studies focus on the diagnosis of COVID-19 patients using serological tests. To detect and assess asymptomatic infections of COVID-19 among people in Wuhan, the epicenter of the COVID-19 pandemic in China, and provide evidence for planning adequate public health measures, we collected and analyzed the clinical data in the Wuhan General Hospital mandatory for 16- to 64-year-old asymptomatic people. This retrospective study estimated the seroprevalence of IgM and IgG and compared the epidemiological characteristics of asymptomatic SARS-CoV-2-infected population. Methods Demographical and radiological data were collected from the Wuhan General Hospital between March 26 and April 28, 2020. Serological tests for IgM and IgG antibodies against SARS-CoV-2 were conducted with a colloidal gold method. Nucleic acid sequences of viruses were detected with RT-PCR. Statistical analyses were carried out using SPSS 20.0 software. Findings Between March 26 and April 28, 2020, 18,391 asymptomatic back-to-work participants were enrolled. Among them, 89 had positivity for IgM (0.48%, 95% confidence interval (CI): 0.38-0.58%); 620 cases had IgG positivity (3.37%, 95% CI: 3.11-3.64%), and 650 cases had either IgG positivity or IgM positivity (3.53%, 95% CI: 3.26-3.80%). After standardizing for the genders and ages in the population of Wuhan, the overall standardized seroprevalence of IgG was 3.33% (95% CI: 3.07-3.59%) and the standardized seroprevalence of IgG was 3.01% (95% CI: 2.69-3.33%) among males and 3.66% (95 % CI: 3.23-4.09%) among females. The standardized seroprevalence of IgG was higher in women than in men with a significant difference (χ2 = 2,060.3, p < 0.01). By a detection method adjustment, the seroprevalence of IgG was 1.57% (95% CI: 1.39-1.75%) in all medical records, of which males were 1.96% (95% CI: 1.64-2.28%), and females were 1.19% (95% CI: 0.99-1.39%). The assay-adjusted seroprevalence of IgG was higher in women than in men, and the difference was significant (χ2 = 5,871.0, p < 0.01). The differences were significant for the seroprevalence of IgG among people who went back to work in different categories of workplace (χ2 = 198.44, p < 0.01). The differences in seroprevalence for IgG positivity or IgM positivity among people who went back to work in different urban and rural areas was also significant (χ2 = 45.110, p < 0.01). Calculated as IgG and/or IgM antibody positivity, the number of new infections was reduced by 64.8% from March 26 to April 28, 2020. Based on the census population aged 16-64 years in Wuhan in 2017, we estimated that 172,340 (95% CI: 157,568-187,112) asymptomatic people aged 16-64 years were infected with SARS-CoV-2 in Wuhan between March 25 and April 28, 2020. This estimate was 3.4-times higher than the officially reported 50,333 infections on April 28. Interpretation The seropositivity rate in Wuhan indicated that RT-PCR-confirmed patients only represented a small part of the total number of cases. Seropositivity progressively decreased in the Wuhan population from March 26 to April 28, 2020, comparable to Japan and Denmark, but well below the level reported in New York, Iran, Italy, and Germany. The prevalence of asymptomatic infection was higher in women than in men among people who went back to work in Wuhan. The low seroprevalence suggests that most of the population remains susceptible to COVID-19. Funding The Emergency Management Project of the National Natural Science Foundation of China (81842035) and Advisory Research Project of the Chinese Academy of Engineering in 2019 (2019-XZ-70).