Abstract Prefrontal cortex (PFC) plays a pivotal role in distributed neural networks orchestrating human behavior. Notably, PFC dysfunction is observed in numerous debilitating developmental, neurological and psychiatric disorders. This emphasizes the need for improved knowledge of the physiology implementing goal-directed behavior to better understand these devastating neuropsychological impairments. The last two decades have seen an exponential increase of PFC research in animals and humans but numerous questions remain. Intracranial EEG recording (iEEG) in humans provides a powerful method to assess the spatial-temporal structure of network activity enabling imaging of both local and inter-areal dynamics. Local neural activation is measured with high frequency activity (HFA; 70-200 Hz) and single-unit activity (SUA). Inter-areal interactions are assessed using connectivity metrics including phase slope index, directional phase amplitude coupling, Granger causality, mutual information and single-trial HFA latency onsets. We combine these local and network approaches with iEEG recordings to address four core areas of PFC dependent function including attention, working memory, learning and interpersonal communication. Aim 1 draws on our prior iEEG work documenting lateral PFC-hippocampal theta band engagement in working memory (WM) and rhythmic PFC- Parietal attentional sampling in the theta range. Here we test the hypothesis that shared and distinct theta networks support WM and attention. Critically, we record from the same individual utilizing identical stimuli differing only in task requirements. In a second study, we employ a naturalistic slot machine gambling task positing that orbital PFC-insular- anterior cingulate interactions track wheel-spinning expectations determining both satisfaction and the motivation to spin again. The slot machine expectation manipulation we employ is banned in casinos given its powerful drive to continue gambling. Aim 2 assesses feedback- based learning in two studies. One posits lateral PFC-hippocampal interactions in rule-based concept learning, the other proposes a central role for orbital PFC-hippocampal interactions in flexibly responding to changing outcome contingencies. Aim 3 proposes novel dual iEEG studies examining how interacting humans exchange knowledge. One task requires transferring knowledge about items and a second examines non-verbal social communication. Dual iEEG involves recording from two interacting patients and preliminary data indicates synchrony between patients in PFC-hippocampal-anterior cingulate circuits during knowledge exchange. Taken together, the proposed work will advance our understanding of how PFC dependent networks support goal-directed human behavior.