The COVID Stress Scales (CSS; Taylor et al., 2020a) were developed as a measure of stress related to any pandemic, making it a valuable self-report measure for potential future large-scale infectious disease outbreaks. Although the initial validation study of the CSS indicated evidence of its psychometric strength, further assessment is warranted of the measure’s retest reliability, the long-term stability of its five-factor structure, discriminant validity, and its relationship with pandemic-related risk and precautionary behaviors. We examined these psychometric elements of the CSS using longitudinal data collected from a nationally representative sample of adults (N = 205) during the COVID-19 pandemic. Results indicate moderate retest reliability that is consistent with changing rates of morbidity. Results also support the five-factor structure and demonstrate convergent validity, discriminant validity, and the ability to predict relevant behavioral responses to an active pandemic. Use of the CSS for future pandemics is discussed.