Thin bamboo wall may easily be penetrated through when installing the common type TDP sensor of 20 mm long for sap flow measurement. Differing with the case of thermal conduction in wood, the part of the upper probe penetrating into the bamboo cavity dissipates heat in a way of radiation. Hence, the dissipation, together with those unevenly distributed waterconducting units within the bamboo wall, may result in measurement inaccuracy of sap flux density and shorter probes have been recommended in general. However, lacking of in situ validation of shorter probes limits their applications for accurate sap flow measurements. In this study, sap flow of Bambusa chungii was simultaneously monitored using selfmade 5, 8 and 10 mm long TDPs in the field for testing their applicability as well as providing technical supports after theoretical analyses. Experimental results indicated that the difference between the fitted equation for calculating sap flux density that was established for B. chungii and the empirical one that was originally proposed by Granier was mainly due to the significant deviation of the parameter α. Such a discrepancy may be preliminarily resulted from the heat accumulating effect in the bamboo cavity that causes higher temperature in the area around the upper probe as well as the lower thermal diffusion coefficient (h₀) of bamboo wall that, with a low water content, may also hoard heat and thus generate larger temperature difference between upper and lower probes. In addition, the way to insert probes into the bamboo wall also affected the measurement and might be another important factor leading to significant variation in monitored sap flux density of B. chungii even with TDPs at the same length. Further analyses on h₀ of surrounding media of TDP provided reasonable explanation for these discrepancies.