Abstract: Abstract: Soil thermal properties, including volumetric heat capacity, thermal diffusivity, and thermal conductivity, are basic physical parameters for determining the change rate of soil temperature, heat storage and transfer. The heat pulse technique, with the advantages of relative easy operation, minimal soil disturbance, and making repeated and automatic readings, has been used widely for measuring in-situ soil thermal properties. A heat pulse is emitted from a line source enclosed in a stainless heating needle and the temperature rises with time at a shorter distance from the heater are recorded for a few minutes. Soil thermal properties are then estimated from the temperature change by time data. For simplicity, the heat pulse probe is normally considered as a line source with infinitesimal probe radius and zero heat capacity when soil thermal properties are calculated. In reality, the finite properties of the probe itself, including finite heat capacity and finite probe radius, can lead to biased thermal property estimations. In this study, we compared the results of soil thermal property estimations with the PILS (pulsed-infinite-line-source) theory and ICPC (identical cylindrical perfect conductors) theory, to evaluate the influences of finite properties of the probe on soil thermal property estimations. The heat pulse probe consist of 3 needles with a diameter of 2 mm and a length of 40 mm. Heat pulse measurements were conducted on a sand soil with water content varied from air dry condition to field capacity, and soil heat capacity, thermal diffusivity, and thermal conductivity were estimated with both the PILS and ICPC methods. In addition, heat capacity estimates with the de Vries model were used to evaluate the accuracy of heat capacity measurements. The results indicated that compared with the PILS theory, the ICPC solution significantly reduced the errors in soil thermal property estimations from the temperature change-by-time curves. For water content ranging from 0.03 to 0.25 m3/m3, the PILS theory underestimated soil thermal conductivity and thermal diffusivity by 11.8% and 5.2%, respectively. Compared with the theoretical values from the de Vries model, the PILS theory and the ICPC theory overestimated soil heat capacity by 16.1% and 7.9%, respectively. Further analysis showed that that the influences of finite probe properties on thermal property estimations were most significant on dry samples, and the errors were reduced linearly with increasing soil water content. The experimental results from this study support that the theoretical analysis including finite heat capacity and finite probe radius improves the accuracies of soil thermal property estimations. The conclusions also have implications in optimizing the design of heat-pulse probes, especially for probes with relatively larger diameters.