Multiple studies have demonstrated that racial/ethnic (R/E) minority communities in US urban areas were disproportionately affected in the early period of COVID-19 pandemic. As the pandemic expanded into rural areas, the degree to which these disparities persist and the role of community-level determinants of health remains unclear. In this longitudinal analysis, monthly COVID-19 infections and deaths were obtained for large metropolitan, medium/small metropolitan, and non-metropolitan counties (NCHS designation) through September 2020 and modeled using separate negative binomial mixed effects models with random state- and county-level intercepts and non-linear trends. Counties were categorized into quartiles based on the percent of R/E minorities and models were adjusted for several county-level sociodemographic variables. Counties in the highest quartile of R/E minorities experienced significantly higher infection and death rates as compared with counties with the greatest share of white residents. This persisted across geographies for nearly the duration of the analysis period, although the magnitude of the difference was greatest in non-metropolitan counties in summer months.