BACKGROUND: Off-target central nervous system (CNS) effects are associated with androgen receptor (AR)-targeting treatments for prostate cancer. Darolutamide is a structurally distinct AR inhibitor with low blood-brain barrier penetration. OBJECTIVE: We compared cerebral blood flow (CBF) in grey matter and specific regions related to cognition after darolutamide, enzalutamide, or placebo administration, using arterial spin-label magnetic resonance imaging (ASL-MRI). METHODS: This phase I, randomized, placebo-controlled, three-period crossover study administered single doses of darolutamide, enzalutamide, or placebo to 23 healthy males (aged 18-45 years) at 6-week intervals. ASL-MRI mapped CBF 4 h post-treatment. Treatments were compared using paired t-tests. RESULTS: Drug concentrations during scans confirmed similar unbound exposure of darolutamide and enzalutamide, with complete washout between treatments. A significant localized 5.2% (p = 0.01) and 5.9% (p < 0.001) CBF reduction in the temporo-occipital cortices was observed for enzalutamide versus placebo and versus darolutamide, respectively, with no significant differences for darolutamide versus placebo. Enzalutamide reduced CBF in all prespecified regions, with significant reductions versus placebo (3.9%, p = 0.045) and versus darolutamide (4.4%, p = 0.037) in the left and right dorsolateral prefrontal cortices, respectively. Darolutamide showed minimal changes in CBF versus placebo in cognition-relevant regions. CONCLUSIONS: Darolutamide did not significantly alter CBF, consistent with its low blood-brain barrier penetration and low risk of CNS-related adverse events. A significant reduction in CBF was observed with enzalutamide. These results may be relevant to cognitive function with early and extended use of second-generation AR inhibitors, and warrant further investigation in patients with prostate cancer. TRIAL REGISTRATION NUMBER: NCT03704519; date of registration: October 2018.