Background and Objective: Epilepsy affects functional connectivity (FC), yet post-operative shifts in FC have rarely been assessed to understand network-level changes in those undergoing lobectomy. We introduce a novel approach that not only assesses whether networks become more or less correlated after resection but also tracks directionality changes in FC on a parcel-parcel, pre-/post-resection basis to characterize post-lobectomy patients. Methods: Pre- and post-lobectomy tractography and resting-state fMRI series were acquired. FC matrices were generated for the affected (operated) and unaffected (unoperated) hemisphere, excluding parcels within resection boundaries. Post-operatively, each connection was binned by directionality. We thus assessed the “fate” of each parcel-to-parcel connection by comparing its pre-operative and post-operative FC. Connections were classified as hypercoupled if their correlation or anticorrelation became stronger post-operatively, uncoupled if their correlation or anticorrelation became weaker, or switched if the direction of the connection changed from correlated to anticorrelated, or vice versa. A two-way repeated-measures ANOVA was performed to assess directionality distributions between networks and bins followed by post-hoc Tukey multiple comparison tests.Results: Our five patient study comprised frontal, temporal, and occipital lobectomies, with all patients achieving seizure freedom post-operatively. Uncoupling was the most common post-operative “fate” (χ²(2,n=80)=79.89, p<0.0001), compared to hypercoupling and directionality switch. This relationship was driven primarily by networks in the affected hemisphere. Bin contribution to post-operative directionality differed significantly across networks (F(1.993,63.78)=82.82, p<0.0001). Anomaly analysis demonstrated an overall reduction in hyperconnected anomalies post-operatively, regardless of lobectomy location.Conclusion: In this five-patient pilot study, epilepsy lobectomy was associated with reduced post-operative FC magnitude and hyperconnected anomaly burden in residual cortical networks. These findings support further prospective study of post-operative network quiescence as a candidate imaging biomarker, though small sample size and heterogeneous imaging timing limit predictive conclusions.