Alcohol Use Disorder (AUD) adversely affects the lives of millions of people, but still lacks effective treatment options. Recent advancements in psychedelic research suggest psilocybin to be potentially efficacious for AUD. However, major knowledge gaps remain regarding (1) psilocybin’s general mode of action and (2) AUD-specific alterations of responsivity to psilocybin treatment in the brain that are crucial for treatment development. Here, we conducted a randomized, placebo-controlled crossover pharmaco-fMRI study on psilocybin effects using a translational approach with healthy rats and a rat model of alcohol relapse. Psilocybin effects were quantified with resting-state functional connectivity using data-driven whole-brain global brain connectivity, network-based statistics, graph theory, hypothesis-driven Default Mode Network (DMN)-specific connectivity, and entropy analyses. Results demonstrate that psilocybin induced an acute wide-spread decrease in different functional connectivity domains together with a distinct increase of connectivity between serotonergic core regions and cortical areas. We could further provide translational evidence for psilocybin-induced DMN hypoconnectivity reported in humans. Psilocybin showed an AUD-specific blunting of DMN hypoconnectivity, which strongly correlated to the alcohol relapse intensity and was mainly driven by medial prefrontal regions. In conclusion, our results provide translational validity for acute psilocybin-induced neural effects in the rodent brain. Furthermore, alcohol relapse severity was negatively correlated with neural responsivity to psilocybin treatment. Our data suggest that a clinical standard dose of psilocybin may not be sufficient to treat severe AUD cases; a finding that should be considered for future clinical trials.