Imbedded thin‐film thermocouples can be used for the measurement of transient surface temperatures with high spatial resolution and sensitivity. To minimize environmental effects it is desirable to have close proximity between the reference junction and the sensor junction. In this article, we present and compare two calibration techniques for such thermocouples. The first technique is a steady‐state technique and uses a focused laser beam with a diameter of 3 &mgr;m as a point heat source. This technique consists of two steps. The first step is an absolute, but spatially averaged temperature measurement with a resistive temperature sensor. In the second step, the uncalibrated thermocouple junction is translated to resolve the spatial temperature profile generated by the laser spot. Both measurements are then combined to provide an absolute calibration which is independent of the thermal properties of the imbedding material and of the absorbed laser power. We used a simple sensor design which consists of a Ni resistive stripe, and the thermocouple junctions it forms with Ta/Au/Ta electrical leads. The second calibration technique is based on a dynamic measurement. To implement this technique we fabricated a small electrical heating element close to one of the junctions. With the dynamic calibration, the time‐resolved response of the thermocouple to a short heat pulse is recorded. A least‐squares fit of a simple model of heat diffusion to the measured response yields both the temperature calibration of the thermocouple and the heat diffusivity of the imbedding medium. The temperature calibration of the thermocouple obtained with the dynamic technique is in good agreement with the steady‐state calibration.