The crosslinking homogeneity of injection molded rubber parts is strongly influenced by the initial bulk temperature of the rubber compound. Therefore, the homogenization of bulk temperature should be aspired in order to improve part quality. To realize this, it is important to know the actual axial bulk temperature distribution in the injection chamber. In this work, a new temperature measurement system was developed. It consists of a movable lance featuring an embedded fast reacting thermocouple. To be able to position this lance at the exact measurement points in the injection chamber of the rubber injection unit, it is mounted on the moving half of the clamping unit and guided into the injection chamber. In order to evaluate the measurement system, it was calibrated and then the response behavior was characterized. The bulk temperature of the rubber compound could be determined with an accuracy of ± 0.1 K and the response behavior of the lance provided to be extremely fast in comparison to other evaluated temperature measurement systems. A factorial design of experiments was developed and implemented, to evaluate the influence of the process conditions on the axial temperature profile in the injection chamber. The factors screw speed, injection chamber temperature, cylinder temperature, cycle time and the viscosity of the rubber compound were varied to characterize their influence on both the bulk temperature and the bulk temperature distribution. Thereby, the screw speed, cycle time and the material used showed a strong impact on the axial temperature distribution in the injection chamber of a state of the art FIFO-rubber injection unit.