Root-phosphorus-uptake (RPU) is a very important parameter for farmland fertilizer management and non-point source pollution control. Since it is almost impossible to determine RPU rate distribution in soil profile directly, it is often estimated using an inverse method, which, however, is influenced by factors, such as the spatial interval (SI) of soil soluble phosphorus (P) concentration measurements, the time interval (T) between the two successive measurements, and the measurement errors (ε_pa) of soil soluble P concentration, etc. In this study, several numerical simulation experiments were first conducted to explore effect of SI, T andε_pa on accuracy and stability of the method. Results show that the overall relative error between the estimated and the theoretical average root-phosphorus-uptake rate distributions was less than 10% when (a) SI was set as 5 cm within the 0~30 cm soil layer and 10~20 cm in soil layers beyond 30 cm in depth; (b) T was set between 5 d and 20 d; and (c) ε_pa was controlled no more than 15%. On such a basis, the inverse method was applied to estimating average root-phosphorus-uptake rate distribution of winter wheat at the seedling stage in a soil column experiment. The estimated total P uptakes by winter wheat at various growth stages matched well with the measured values, with all the relative errors ranging within 10%.