Abstract Background Goal-directed movements involve integrating proprioceptive and visuo-motor information. Although the neural correlates of such information processing are known, the details of how sensory-motor integration occurs are still largely unknown. Objective The study aims to characterize movements with different sensory goals, by contrasting the neural activity involved in processing proprioceptive and visuo-motor information. To accomplish this, we have developed a new methodology that utilizes the irregularity of the instantaneous gamma frequency parameter for characterization. Approach In this study, 8 essential tremor patients undergoing an awake deep brain stimulation (DBS) implantation surgery repetitively touched the clinician’s finger (forward visually-guided/FV movement) and then one’s own chin (backward proprioceptively-guided/BP movement). Neural electrocorticographic (ECoG) recordings from the motor (M1), somatosensory (S1), and posterior parietal cortex (PPC) were obtained and band-pass filtered in the gamma range (30-80Hz). The irregularity of the inter-event intervals (IEI; inverse of instantaneous gamma frequency) were examined as: 1) correlation between the amplitude and its proceeding IEI, and 2) auto-information of the IEI time series. We further explored the network connectivity after segmenting the FV and BP movements by periods of accelerating and decelerating forces, and applying the IEI parameter to transfer entropy methods. Results Conceptualizing that the irregularity in IEI reflects active new information processing, we found the highest irregularity in M1 during BP movement, highest in PPC during FV movement, and the lowest during rest at all sites. Also, connectivity was the strongest from S1 to M1 and from S1 to PPC during FV movement with accelerating force and weakest during rest. Significance We introduce a novel methodology that utilize the instantaneous gamma frequency (i.e., IEI) parameter in characterizing goal-oriented movements with different sensory goals, and demonstrate its use to inform the directional connectivity within the motor cortical network. This method successfully characterizes different movement types, while providing interpretations to the sensory-motor integration processes.