OBJECTIVES: To determine SARS-CoV-2 seroprevalence over time and risk factors among pregnant women at delivery in São Paulo, Brazil; and to evaluate the suitability of pregnant women as a sentinel population for SARS-CoV-2 serosurveillance. METHODS: Unselected consecutive pregnant women presenting at the labor ward of a single large hospital between July 20th 2020 to February 21st 2021 were enrolled and tested for SARS-CoV-2 serology using two assays: the rapid chromatic Wondfo One Step (for total IgA and IgG detection) and Roche Elecsys assay (detecting anti-nucleoprotein [N] IgG). SARS-CoV-2 seroprevalence was computed as smooth spline function over time with 95% confidence intervals (CI). Risk factors were evaluated for positivity by each assay. We compared timepoint seroprevalence by the two assays with four concomitant community household surveys (HHS), in which the Roche assay was used, to determine the sensitivity and relevance of the pregnant women population as sentinel population. RESULTS: Overall SARS-CoV-2 seroprevalence was 28.9% (221/763) by Roche and 17.9% (137/763) by Wondfo. Reported symptoms experienced during pregnancy were all significantly correlated with being SARS-CoV-2 seropositive at delivery with any assay (with odds-ratios ranging from 3.0 [95% CI: 2.1-4.3] for coryza to 22.8 [95% CI: 12.3-46.6] for ageusia). Seropositivity by either assay was high in women at delivery in the early period of the pandemic (June 2020), compared with seropositivity in women from the concomitant HHS: 44.1% (95% CI: 21.8-66.4) for Roche, 54.1% (30.9-78.5) for Wondfo, versus 11.4% (95% CI: 9.2-13.6) for HHS. For later periods (October 2020 and January 2021), the seropositivity in women at delivery measured by Roche corresponded well with the prevalence found among women in the HHS using the same assay, whilst prevalence measured by Wondfo dropped. CONCLUSIONS: Women at delivery represent a highly exposed and readily accessible population for sentinel surveillance of emerging infections such as SARS-CoV-2.