Although previous research has found a high prevalence of anxiety during the COVID-19 pandemic among healthcare workers, longitudinal studies on post-pandemic anxiety and predictor variables have been less abundant. To examine the evolution of anxiety in healthcare workers from the beginning of the pandemic until one and a half years later, analyzing the influence of occupational and psychosocial variables, as well as their possible predictors. : This was a prospective longitudinal design with three periods of data collection: (1) between 5 May and 21 June 2020, (2) six months after the end of the state of alarm (January-March 2021), and (3) one year after this second assessment (April-July 2022), in which generalized anxiety (GAD-7) was evaluated, as well as occupational and psycho-emotional variables (i.e., social support, self-efficacy, resilience, and cognitive fusion) in healthcare workers in direct contact with COVID-19 patients in Spain. : A high prevalence of anxiety was found, with a clear decrease over time. Associations were found between anxiety and certain sociodemographic and work variables (i.e., years of experience, 0.046; COVID-19 symptoms, 0.001; availability of PPE, 0.002; workload, 0.001; family contagion concern, 0.009). Anxiety maintained negative relationships with social support ( 0.001), self-efficacy ( 0.001), and resilience ( 0.001) and positive associations with cognitive fusion ( 0.001). Cognitive fusion seemed to be a clear predictor of anxiety. : Our findings suggest that social support, self-efficacy, and resilience act as buffers for anxiety, whilst cognitive fusion was found to be a clear risk factor for anxiety. It is important to emphasize the risk role played by cognitive fusion on HCWs as a clear risk factor for stressful work events. The findings emphasize the need to implement specific interventions to promote the mental well-being of healthcare workers, particularly in crisis contexts such as the COVID-19 pandemic.