In the multisensory environment, inputs to each sensory modality are rarely independent. Sounds often follow a visible action or event. Here we present behaviorally relevant evidence from the human EEG that visual input prepares the auditory system for subsequent auditory processing by resetting the phase of neuronal oscillatory activity in auditory cortex. Subjects performed a simple auditory frequency discrimination task using paired but asynchronous auditory and visual stimuli. Auditory cortex activity was modeled from the scalp-recorded EEG using spatiotemporal dipole source analysis. Phase resetting activity was assessed using time-frequency analysis of the source waveforms. Significant cross-modal phase resetting was observed in auditory cortex at low alpha frequencies (8-10 Hz) peaking 80 ms after auditory onset, at high alpha frequencies (10-12 Hz) peaking at 88 ms, and at high theta frequencies (∼ 7 Hz) peaking at 156 ms. Importantly, significant effects were only evident when visual input preceded auditory by between 30 and 75 ms. Behaviorally, cross-modal phase resetting accounted for 18% of the variability in response speed in the auditory task, with stronger resetting overall leading to significantly faster responses. A direct link was thus shown between visual-induced modulations of auditory cortex activity and performance in an auditory task. The results are consistent with a model in which the efficiency of auditory processing is improved when natural associations between visual and auditory inputs allow one input to reliably predict the next.