ABSTRACT We present a procedure dedicated to the calibration of a scanning thermal microscopy probe operated in an active mode and a modulated regime especially for the measurement of solid material thermal conductivities. The probe used is a microthermocouple wire mounted on a quartz tuning fork. Measurements on reference samples are performed successively in vacuum and ambient air conditions revealing a clear difference in the dependence of the thermal interaction between the probe and the sample on the sample properties. Analytical modeling based on the resolution of the heat equation in the wire probe and a description of the thermal interaction using a network of thermal conductances are used to fit experimental data and analyze this difference. We have experimentally verified that the effective thermal contact radius of the probe tip depends on the sample thermal conductivity in ambient conditions, whereas it remains constant in vacuum. We have defined the measurement range of the technique based on the decrease in the probe sensitivity at high thermal conductivities. Considering the experimental noise of our electrical device, it is shown that the maximum measurable value of thermal conductivity is near 23Wm−1 K−1 in vacuum and 37 W m−1 K−1 in ambient air conditions. Moreover, the lowest uncertainties are obtained for thermal conductivities below 5Wm−1 K−1 typically.