Background: Research has shown the substantial impact of the COVID-19 pandemic on healthcare workers’ (HCWs) mental health, however, it mostly relies on data collected during the early stages of COVID-19. The aim of this study is to assess the long-term trajectory of HCWs’ mental health and the associated risk factors. Methods: a longitudinal cohort study was carried out in an Italian hospital. At Time 1 (July 2020–July 2021), 990 HCWs took part in the study and completed the General Health Questionnaire (GHQ-12), the Impact of Event Scale (IES-R), and the General Anxiety Disorder (GAD-7)questionnaire. McNemar’s test measured changes in symptoms’ trajectories, and random effects models evaluated risk factors associated with scores above the cut-off. Results: 310 HCWs participated to the follow-up evaluation (Time 2; July 2021–July 2022). At Time 2, scores above cut-offs were significantly lower (p < 0.001) than at Time 1 for all scales (23% vs. 48% for GHQ-12; 11% vs. 25% for IES-R; 15% vs. 23% for GAD-7). Risk factors for psychological impairment were being a nurse (IES-R: OR 4.72, 95% CI 1.71–13.0; GAD-7: OR 2.82, 95% CI 1.44–7.17), a health assistant (IES-R: OR 6.76, 95% CI 1.30–35.1), or having had an infected family member (GHQ-12: OR 1.95, 95% CI 1.01–3.83). Compared to Time 1, gender and experience in COVID-19 units lost significance with psychological symptoms. Conclusions: data over more than 24 months from the pandemic onset showed improvement of HCWs’ mental health; our findings suggested the need to tailor and prioritize preventive actions towards healthcare workforce.