In this study, we used a in-silico cardiac wedge model with high spatial resolution to evaluate the role of transmural heterogeneity on T-wave morphology. Computer simulations were performed with transmural electrograms recorded with a pacing rate of 60 bpm. QT and TpTe intervals were extracted. The model consisted of i) a human left-ventricular wedge mesh (∼4 millions cells); and ii) modified version of the ten Tuscher 2006 cardiac cell model with an added late sodium current. Heterogeneity was generated by changing the spatial distribution of cell types, i.e., M-cells and Epi-cells, across the wedge. Three different cell distribution were used in the simulations: i) random; ii) layers; and iii) layered then gradient. The wedge was stimulated to produce a normal activation propagating from endocardium to epicardium. Our simulations showed that a positive T-wave required the repolarization wave to predominately propagate in opposite direction to the depolarization wave with longer APDs in the endocardial region compared to epicardium. Moreover, QT intervals increased with the amount of Mcells in the wedge model while TpTe intervals were dependent on the transmural heterogeneity. In summary, we showed that cell distributions highly affect both repolarization and T-wave morphology parameters in a high-resolution human wedge model.