Abstract: Abstract: Accurate measurement of near-surface soil water content (0-5 cm) is essential for studying water and heat exchange between land and atmosphere. Heat-pulse technique has potential to measure soil volumetric water content, but only used in laboratory, greenhouse and field at depths below 5 cm. The objective of this study was to test the possibility of heat-pulse technique for measuring water content in near-surface soil layer (0-5 cm) in the field. The experiment was conducted in a sandy loam soil in an experimental plot (2 m×2 m×2 m) in China Agricultural University. Multi-needle heat-pulse sensor was used to measure volumetric heat capacity in various depths above 5 cm and four multi-needle heat-pulse sensors were used. Average volumetric heat capacity from four multi-needle heat-pulse sensors was calculated at each depth and the values at four depths (3, 9, 21 and 39 mm) were extracted to be analyzed. When calculating soil volumetric heat capacity from temperature change versus time data, the pulse infinite line source theory was used at 9, 21 and 39 mm, while for volumetric heat capacity at 3 mm, the pulse infinite line source-adiabatic boundary condition was used because of the soil-air interface effect. Based on the linear relationship of soil volumetric heat capacity and volumetric water content in mineral soil, soil volumetric water content could be calculated from soil volumetric heat capacity, volumetric heat capacity of water, soil bulk density and solid heat capacity, which was named heat-pulse water content method. When the temporal series of soil volumetric heat capacity were measured, the change in soil water content could be calculated without knowledge of soil-specific properties. With the initial water content measured by another independent method, dynamic of soil water content could be determined from the change in soil water content series data and this method was called heat-pulse water content change method. Water contents at those four depths were calculated by the above two methods respectively. Besides, a ring device was made for collecting and splitting soil samples into nine layers in 0- to 5-cm depth. With the ring device, water contents at 3, 9, 21 and 39 mm were determined by the traditional gravimetric method. Water contents from gravimetric method were used to evaluate accuracy of water contents from heat-pulse technique. The results showed that compared with soil water contents from gravimetric method, root mean square errors of soil water contents from heat-pulse water content method were 0.082, 0.075, 0.018 and 0.018 m3/m3 and mean deviations were 0.076, 0.073, 0.009 and 0.010 m3/m3 at four depths, while for the values from heat-pulse water content change method, root mean square errors were 0.022, 0.006, 0.004 and 0.006 m3/m3 and mean deviations were 0.048, 0.024, 0.017 and 0.025 m3/m3, indicating that heat-pulse water content change method could obtain more accurate results. The maximum standard deviations of water contents from heat-pulse water content method were 0.077, 0.077, 0.044 and 0.059 m3/m3 at four depths. For soil water contents from heat-pulse water content change method, the maximum standard deviations were 0.061, 0.052, 0.019 and 0.021 m3/m3, all smaller than those values from heat-pulse water content method, showing heat-pulse water content change method decreased the differences of water content measurements among four multi-needle heat-pulse sensors. The results showed that heat-pulse water content change method could measure soil water content accurately at depths of 3, 9, 21 and 39 mm, with root mean square error less than 0.022 m3/m3. Heat-pulse technique can accurately measure water content in near-surface soil layer, but further research is needed for accurate water content measurement in 0-3 mm soil layer.