Background Accurate estimates of the prevalence of infections play an important role in COVID-19 surveillance. Older people are known to have higher risks of severe outcomes after infection, but whether they also have a higher infection rate remains unclear. To obtain estimates of COVID-19 prevalence among older people, we synthesized evidence from RT-PCR-based prevalence and serological studies. Methods We conducted a scoping review using a comprehensive search of MEDLINE (Ovid), Embase (Ovid), Europe PMC, ClinicalTrials.gov, and the WHO COVID-19 Research Database from December 2019 to Oct 2022. We included population-based cross-sectional (sero)prevalence studies among older people (i.e., people aged >= 65 +/- 5 years) who were tested for SARS-CoV-2 infection using RT-PCR tests, antigen tests, or serological tests. Studies that were conducted solely in institutional housing were excluded. Eligible studies were extracted and critically appraised. We described and mapped the prevalence (tested by RT-PCR or antigen tests) and seroprevalence (tested by serological tests) by geographical area and time. We then compared the estimated prevalence with WHO-reported prevalence and the prevalence among younger age groups from the same study. Results We identified 202 (sero)prevalence estimates from 126 studies, covering 50 countries up to October 2022. Of the 126 studies, 28 estimated RT-PCR-based prevalence; 104 studies estimated seroprevalence, ranging from 0% in Jordan to 22.5% in the United States in 2020, from 0.41% in Brazil to 98% in Chile in 2021. In the year 2020, the prevalence of COVID-19 ranged from 0.0006% in China to 52.8% in Brazil, while in 2021, the prevalence ranged from 0.06% in England to 41.1% in Brazil. The ratio of the reported prevalence to estimated prevalence ranged from <0.01 to 77.50, where 86% (24/28) studies estimated a higher prevalence than WHO reported, and half of them estimated >10 times higher prevalence. One-third of studies (32%, 9/28) estimated a higher prevalence in older people than younger people. Conclusions Our findings suggest that underreporting of COVID-19 cases among older people may exist extensively worldwide. Compared with younger groups, older people were less likely to be infected with COVID-19 in two-thirds of the studies through the first two years of the pandemic.