BACKGROUND: Post-acute sequelae of SARS-CoV-2 infection (PASC) are ongoing, relapsing, or new symptoms present at least 3 months after infection. Predictors of PASC, particularly across diverse racial and ethnic groups, remain unclear. OBJECTIVES: Assess the prevalence of PASC 1 year after infection, examining differences in PASC prevalence by the social construct of race. DESIGN: Retrospective observational cohort study. PARTICIPANTS: In total, 863 adults aged 18 years or older, who were assessed for COVID-19 at two hospitals in New York City during the first two waves of the pandemic (March to July 2020 and January to March 2021). MAIN MEASURES: Prevalence of self-reported PASC, including its respiratory and neurological phenotypes. Data were gathered via telephone surveys conducted 11-15 months following diagnosis. Logistic regression models were used to identify predictors of PASC. KEY RESULTS: In total, 54.1% of those diagnosed with COVID-19 reported PASC symptoms 1 year after infection. Racial and ethnic disparities in the prevalence of PASC varied by PASC phenotype and by the time of initial diagnosis (Wave 1 versus Wave 2). Asian adults had significantly lower odds of reporting any PASC compared to White adults (AOR = 0.55, p=0.02), particularly for neurological symptoms (AOR = 0.5, p=0.01). Black adults had significantly higher odds of reporting respiratory PASC (AOR = 2.67, p<0.001) and lower odds of neurological PASC (AOR = 0.54, p=0.02). Females had higher odds of respiratory (AOR = 1.45, p=0.04) and neurological PASC (AOR = 1.45, p=0.02). Loneliness was consistently associated with higher odds of all PASC categories. CONCLUSIONS: This study reveals a high prevalence of PASC 1 year post-infection, with notable racial and ethnic disparities. The results underscore the need for long-term monitoring of those infected with COVID-19 during the initial waves, with a focus on identifying and addressing yet unmeasured social determinants of health that contribute to these disparities.