Until recently, the incidence of COVID-19 was primarily estimated using molecular diagnostic methods. However, the number of cases is vastly underreported using these methods. Seroprevalence studies estimate cumulative infection incidences and allow monitoring of transmission dynamics, and the presence of neutralizing antibodies in the population. In February 2020, the Mexican Social Security Institute began conducting anonymous unrelated sampling of residual sera from specimens across the country, excluding patients with fever within the previous two weeks and/or patients with an acute respiratory infection. Sampling was carried out weekly and began 17 days before Mexico's first officially confirmed case. The 24,273 sera obtained were analyzed by chemiluminescent-linked immunosorbent assay (CLIA) IgG S1/S2 and, later, positive cases using this technique were also analyzed to determine the rate of neutralization using the enzyme-linked immunosorbent assay (ELISA). We identified 40 CLIA IgG positive cases before the first official report of SARS-CoV-2 infection in Mexico. The national seroprevalence was 3.5% in February and 33.5% in December. Neutralizing activity among IgG positives patients during overall study period was 86.1%. The extent of the SARS-CoV-2 infection in Mexico is 21 times higher than that reported by molecular techniques. Although the general population is still far from achieving herd immunity, epidemiological indicators should be re-estimated based on serological studies of this type.