In rubber injection molding, it is common to use heated clamping plates to temper molds. This often leads to an inhomogeneous cavity temperature profile which results in an unbalanced curing degree. The aim of this work is to evaluate various heating concepts for their suitability to deliver a homogenous mold temperature profile which is needed for uniform part quality. Therefore, a thick-walled hollow tube was selected as test geometry for which different heating concepts were simulated: 1) conventional, indirect heating via clamping plates, 2) improved heat flow to core section of mold by copper insert with high thermal conductivity and 3) active heating of core section with implemented heater cartridge. Practical experiments were executed with the conventional concept and the actively heated core using an NBR test compound. At least 3 wall temperature sensors were used to evaluate the mold temperature field. The part quality was tested by compression set measurements with specimens being extracted on locally different positions. Simulation results correlate very well to the practical experiments. The conventionally heated mold shows local cavity temperature differences up to 40°C which can only be avoided in reducing the set temperature and a significant increase of curing time. While the core with copper insert reduces this effect, only the actively heated core allowed mold temperatures above 150°C maintaining the required temperature homogeneity. Thereby, cycle time reductions of up to 50% compared to the conventional process are conceivable. Finally, a new concept was developed which allows to dynamically control the mold cavity temperature. Ceramic heaters were placed close to the cavity surface, enabling rapid mold heating and acceleration of curing due to a defined temperature increase in the curing phase. Simulations demonstrate that especially peroxide curing materials can be processed far more efficiently compared to an isothermally tempered mold.