OBJECTIVE: This study aimed to explore variations of brain functional connectivity patterns among post-COVID-19 patients with different outcomes of sleep quality. METHODS: Post-COVID-19 patients were prospectively enrolled and categorized into improvement or deterioration groups based on changes in sleep quality after a three-month follow-up. Functional MRI and blood samples were collected, while a battery of assessments was administered to evaluate sleep quality, mental status, and cognition. Baseline and follow-up data were compared to identify post-infection alterations. Brain functional networks and graph theory analysis were employed to derive network properties, with subsequent investigation into the correlation between these properties, sleep and psychological assessment scores, and blood test outcomes. RESULTS: The graph theory analysis revealed a significantly increase in global efficiency (Eglob) and local efficiency (Eloc), and a decrease in λ, in the improvement group. A notable enhancement of frontoparietal network (FPN) were observed. The deterioration group exhibited a significant increase in Eloc and λ, along with a decrease in Eglob. Furthermore, the deterioration group demonstrated a lower level of Eglob at follow-up. With respect to network strength, all networks except FPN showed significantly higher values in the improvement group. Pittsburgh Sleep Quality Index and Self-Rating Anxiety Scale scores differed between two groups. CONCLUSION: Changes in sleep quality following COVID-19 infection are associated with brain functional connectivity patterns. Decreased Eglob is related to worsened sleep quality. The normalized strength of FPN serves as a key indicator for improved sleep quality, while other networks also play roles in regulating sleep quality.