Heat pulse (HP) sensors have been used extensively to measure soil thermal properties and volumetric water content (θ) simultaneously. Probe spacing (R), the distance between heater probe and temperature probe, is routinely calibrated in agar solution before making HP measurements. In reality, R may change due to probe deflection at insertion into soil, causing uncertainties in HP measured volumetric heat capacity (C) and θ. In this paper, we present a simple on-site calibration method for determining in-situ R by using theoretical C values estimated from soil bulk density and θ values of intact soil cores collected near the sensor locations at the end of experiment. The proposed approach was tested on two field soils where C and θ were monitored continuously. Inconsistent results were observed between agar-calibrated R values (Ragar) and R values obtained from in-situ calibration (Rin-situ). Maximum difference between Ragar and Rin-situ reached 20% of Ragar. Compared toTDR measured θ, using Ragar led to root mean square errors (RMSEs) of 0.26 MJ m-3 K-1 in C and 0.06 m3 m-3 in θ on average. For Rin-situ, the RMSE of C was less than 0.16 MJ m-3 K-1 and RMSE of θ was within 0.04 m3 m-3. Thus, the use of in-situ probe spacing calibration improved the accuracies of HP measured C and θ.