Human locomotion induces rhythmic movements of the trunk and head. Vestibular signaling is relayed to multiple regions in the brainstem and cerebellum, and plays an essential role in maintaining head stability. However, how the vestibular-cerebellar network contributes to the rhythmic locomotor pattern in humans is unclear. Transcranial alternating current stimulation (tACS) has been used to investigate the effects of the task-related network between stimulation regions in a phase-dependent manner. Here, we investigated the relationship between the vestibular system and the cerebellum during walking imagery using combined tACS over the left cerebellum and alternating current galvanic vestibular stimulation (AC-GVS). In Experiment 1, we tested the effects of AC-GVS alone at around individual gait stride frequencies. In Experiment 2, we then determined the phase-specificity of combined stimulation at the gait frequency. Combined stimulation was applied at in-phase (0° phase lag) or anti-phase (180° phase lag) between the left vestibular and left cerebellar stimulation, and the sham stimulation. We evaluated the AC-GVS-induced periodic postural response during walking imagery or no-imagery using the peak oscillatory power on the angular velocity signals of the head in both experiments. In Experiment 2, we also examined the phase-locking value (PLV) between the periodic postural responses and the left AC-GVS signals to estimate entrainment of the postural response by AC-GVS. AC-GVS alone induced the periodic postural response in the yaw and roll axes, but no interactions with imagery walking were observed in Experiment 1 (p > 0.05). By contrast, combined in-phase stimulation increased yaw motion (0.345 ± 0.23) compared with sham (-0.044 ± 0.19) and anti-phase stimulation (-0.066 ± 0.18) during imaginary walking (in-phase vs. other conditions, imagery: p < 0.05; no-imagery: p ≥ 0.125). Furthermore, there was a positive correlation between the yaw peak power of actual locomotion and in-phase stimulation in the imagery session (imagery: p = 0.041; no-imagery: p = 0.177). Meanwhile, we found no imagery-dependent effects in roll peak power or PLV, although in-phase stimulation enhanced roll motion and PLV in Experiment 2. These findings suggest that combined stimulation can influence vestibular-cerebellar network activity, and modulate postural control and locomotion systems in a temporally sensitive manner. This novel combined tACS/AC-GVS stimulation approach may advance development of therapeutic applications.