This study assessed the current isothermal oil heating method for thermal management of injection molds and explored the feasibility of variothermal induction heating by both experimental trials and computer modeling using multiple commercial solvers (Ansys Maxwell and Mechanical, Sigmasoft, and Autodesk Moldflow). The temporal evolution and spatial distribution of temperature, injection pressure, and end of fill cavity pressure along with the shrinkage and warpage behavior analyzed for discs made from 30% glass-fiber reinforced PEEK were used as qualifying metrics. Using oil tempering, experimental measurements during start-up in mold open state showed the insert temperatures to be lower than vendor recommendation. The thermal model (with Ansys and Sigmasoft) overestimated the heating rate and the steady state temperature while it underestimated the spatial gradients. The injection molding process model (Moldflow and Sigmasoft) accurately predicted the experimentally determined maximum injection pressure of 92 MPa although the exact pressure progression of the Decoupled IIISM process could not be replicated. Sigmasoft correctly estimated the cavity pressure curve until the maximum pressure while Moldflow calculated lower values. The computed shrinkage values corresponded well with the measured data. However, the maximum warpage was underestimated by Moldflow while overestimated by Sigmasoft. The combination of experimental and modeling work established an acceptable spatial gradient criterion based on the temperature distribution in concert with the warpage data. Experimental measurements on a pancake coil and a one way coupled electromagnetic-heat transfer Ansys model for the induction heating system demonstrated an inhomogeneous temperature distribution with unacceptable spatial gradients. It was further established, using the computer model, that the inhomogeneity could be improved by increasing the diameter of the inductor, changing the direction of the conductors, and adding magnetic intensification. Finally, a meander inductor provided the most uniform temperature distribution meeting the theoretically acceptable temperature gradient criterion. Moldflow also predicted the temperature distribution for all inductor geometries similarly and was used to demonstrate that larger temperature gradients such as one obtained from the pancake coil resulted in unacceptably high warpage which could be controlled within acceptable limits using a meander coil. In conclusion, this thesis recommends a review of the oil heater set-up to improve existing mold thermal management. It provides a reasonably calibrated computer model for future use. The most important outcome is the unique design configuration of an induction coil which predicts uniform temperature distribution thereby demonstrating that the variotherm induction heating can be suitable for injection molding a disc shaped part and thus sets the stage for further prototype improvement.