A new approach that does not require additional hardware for temperature-drift compensation is reviewed. The efficiency of the new approach depends on the repeatability of the input transducer temperature-dependent transfer characteristics. A simple polynomial approximation of the measurand model was used in the application example. To bring the temperature-compensation efficiency to the level limited by the repeatability of the transfer characteristics of the input transducer, advanced fitting methods can be used.