Actual melt temperatures are rarely measured directly when conducting rheological characterisations of materials in non-ambient conditions despite the potentially large influence temperature may have on the theological data. For rheometers that use only a temperature-regulated lower plate, it is likely that the set point and true melt temperatures differ, an effect that becomes significant when characterising melts or suspensions close to phase-change events like crystallisation. This work investigates the magnitude of these effects for a controlled-stress rheometer featuring a temperature-controlled lower plate. The lower plate was fitted with a serrated cover disc that was found to exacerbate temperature deviations from the desired set point. Steady-state radial and vertical temperature profiles within the sample were measured and compared with the predictions of a finite element analysis model. The deviations between set point and measured temperatures were successfully predicted by the simulation for two typical gap heights for a thermoplastic, ceramic paste. The non-ideal heat transfer characteristics were also investigated numerically for a representative polymer system that demonstrated the increased deviations from ideal values for lower thermal conductivity materials.