Decision making relies on dynamic interactions of distributed, primarily frontal brain regions. Extensive evidence from functional magnetic resonance imaging (fMRI) studies indicates that the anterior cingulate (ACC) and the lateral prefrontal cortices (latPFC) are essential nodes subserving cognitive control. However, because of its limited temporal resolution, fMRI cannot accurately reflect the timing and nature of their presumed interplay. The present study combines distributed source modeling of the temporally precise magnetoencephalography (MEG) signal with structural MRI in the form of "brain movies" to: (1) estimate the cortical areas involved in cognitive control ("where"), (2) characterize their temporal sequence ("when"), and (3) quantify the oscillatory dynamics of their neural interactions in real time. Stroop interference was associated with greater event-related theta (4 - 7 Hz) power in the ACC during conflict detection followed by sustained sensitivity to cognitive demands in the ACC and latPFC during integration and response preparation. A phase-locking analysis revealed co-oscillatory interactions between these areas indicating their increased neural synchrony in theta band during conflict-inducing incongruous trials. These results confirm that theta oscillations are fundamental to long-range synchronization needed for integrating top-down influences during cognitive control. MEG reflects neural activity directly, which makes it suitable for pharmacological manipulations in contrast to fMRI that is sensitive to vasoactive confounds. In the present study, healthy social drinkers were given a moderate alcohol dose and placebo in a within-subject design. Acute intoxication attenuated theta power to Stroop conflict and dysregulated co-oscillations between the ACC and latPFC, confirming that alcohol is detrimental to neural synchrony subserving cognitive control. It interferes with goal-directed behavior that may result in deficient self-control, contributing to compulsive drinking. In sum, this method can provide insight into real-time interactions during cognitive processing and can characterize the selective sensitivity to pharmacological challenge across relevant neural networks.