Turbochargers develop away from an auxiliary component—being “off the shelve”—towards an integrated component of the internal combustion engine. Hence, increased attention is paid to the accuracy of the measured turbine and compressor maps. Especially turbine efficiency measurement under engine-relevant operating conditions (pulsed flow) is recently receiving increased attention in the respective research community. Despite various turbine map extrapolation methods, sufficient accuracy of the input test data is indispensable. Accurate experimental data are necessary to achieve high quality extrapolation results, enabling a wide range and precise prediction of turbine behavior under unsteady flow conditions, determined by intermittent operation of the internal combustion engine. The present work describes the first application of a contactless shaft torque measurement technique—based on magnetostriction—to a small automotive turbocharger. The contactless torque measuring system is presented in detail and sensor principle as well as sensor calibration are illustrated. A sensitivity study regarding sensor position influences onto sensor signal proves the robustness and very good repeatability of the system. In the second part of the paper, steady state experimental results from operation on a conventional hot gas test stand over a wide map range are presented. These results are validated against full turbine stage (adiabatic as well as diabatic) CFD results as well as against “cold” efficiency measurements, based on measured inlet and outlet temperatures. The influence and relevance of bearing friction for such measurements is underlined and the improvements on this matter—achieved by direct torque measurement—are demonstrated.