Recent research suggests that short-term exposure to air pollution is associated with an elevated prevalence of respiratory infectious disease. We examine the relationship between the air quality index (AQI) and weekly cases of influenza-like illnesses (ILI) and COVID-19 in the United States. We address potential bias from omitted variables and measurement error with an instrumental variable approach using atmospheric temperature inversions. Unlike other recent studies, we find no relationship between air quality and either ILI or COVID-19 cases.