BACKGROUND: Cognitive dysfunction ("brain fog") is a commonly reported post-COVID-19 symptom. Leveraging data from five general population cohorts across four European countries (Estonia, Iceland, Norway, and Sweden), we assessed long-term prevalence of impaired subjective cognitive function among individuals diagnosed with COVID-19 by acute illness severity. METHODS: The included cohorts consisted of adult participants recruited from March 2020 and followed with self-report measures of cognitive function and past COVID-19 infection (except one cohort consisting of clinically confirmed COVID-19 cases) through February 2023. In a cross-sectional analysis we contrasted the prevalence of impaired cognitive function among individuals with and without a COVID-19 diagnosis, overall and by illness severity up to 32 months post-diagnosis. We adjusted for age, gender, education, relationship status, binge drinking, body mass index, previous psychiatric diagnosis, number of chronic medical conditions, and response period. In a longitudinal analysis, we assessed potential changes in cognitive function scores before and after COVID-19 diagnosis. RESULTS: The study population consisted of 153,841 participants (71% women), with 31,359 (20.4%) reporting a positive COVID-19 test. Overall, a COVID-19 diagnosis was not statistically significantly associated with increased prevalence ratio (PR) of impaired cognitive function (PR 1.30 [95% CI: 0.98-1.71]). Individuals bedridden due to COVID-19 for 1-6 days (PR 1.38 [95% CI 0.96-1.99]) or ≥ 7 days (2.59 [1.55-4.33]) had higher prevalence of impaired cognitive function compared to those never diagnosed, while individuals never bedridden had a lower prevalence to those never diagnosed with COVID-19 (0.89 [0.80-1.00]). These findings were corroborated in the longitudinal analysis where a pre- to post diagnosis decline in cognitive function was observed among individuals bedridden due to COVID-19 (p < 0.0001). CONCLUSIONS: The data indicates that a severe COVID-19 acute illness course is associated with impaired cognitive function up to 18-32 months after COVID-19 diagnosis.