Virtual reality exposure therapy (VRET), an innovative form of exposure therapy (ET), has been demonstrated to be effective in treating acrophobia. However, its neural mechanisms of action and how it differs from traditional imaginal exposure therapy (IET) remain unclear. This study utilized resting-state functional magnetic resonance imaging (fMRI) to investigate the effects of VRET on brain activity in acrophobic patients and to explore the potential mechanisms underlying its therapeutic action. Fifty patients with acrophobia were randomly assigned to either an experimental group (25 patients) or a control group (25 patients) based on different treatments. The experimental group received VRET, while the control group received conventional IET. A mixed-design repeated-measures analysis of variance (ANOVA) was performed on the whole brain to identify brain regions affected by the intervention. Both groups of patients underwent treatment twice a week for 3 weeks. fMRI scans were performed for all patients at baseline and after treatment to facilitate a comparison of clinical effects at the end of the treatment period. The degree centrality (DC) values of the blood oxygenation level dependent signals across the entire brain were analyzed. A mixed-design repeated-measures ANOVA was conducted on the pre- and post-intervention data to identify brain regions affected by the intervention. The degree of symptom improvement was assessed using self-report measures, including the Acrophobia Questionnaire (AQ), the Attitude Toward Heights Questionnaire, the Behavior Avoidance Test, and the 7-item Generalized Anxiety Disorder Scale. These assessments were correlated with pre- and post-intervention differences in brain activity. Additionally, a functional connectivity (FC) analysis was conducted to identify any atypical connectivity patterns following the ET. There was a significant positive correlation between the change in scores on the AQ and the right middle temporal gyrus (MTG) ( = 0.442, = 0.045). After VRET, DC values in the right calcarine, MTG, cuneus, and precuneus were decreased ( < 0.005), while DC values in the postcentral gyrus decreased after IET ( < 0.05). Additionally, reduced FC between the right MTG and both the right medial superior frontal gyrus and the left MTG was observed in acrophobia patients following VRET. In the IET group, reduced FC between the left MTG and the left superior temporal gyrus was found ( < 0.005). Preliminary results suggest that VRET may improve abnormal brain activity in acrophobia by modulating the activity of the default mode network and the primary visual cortex.