Both laboratory analysis and pot experiments in greenhouse were conducted by using four typical soils widely distributed in Zhejiang province and rye as testing plant to study the relationship between plant-availability of P and characteristics of phosphate desorption in soils The results obtained demonstrated that P uptake of rye plant was correlated more significantly with desorbed-P than with Olsen-P or Bray 1-P. What is more interesting is that the plant-absorbed P was much more compatible with desorbed-P than with Olsen-P or Bray 1-P in quantity,dinder various levels of P intensity. The Q-l relationship of either P uptake of rye plant or desorbed-P against P intensity fitted much better with the Langmuir equation than that of Olsen-P or Bray 1-P against P intensity as far as the four tested soils are concerned. It suggested that plant-availability of phosphorus in soil was essentially related with phosphate desorption behavior. The maximum buffering capacity of desorption (MBCD) obtained from the Langmuir equation mas found to explain well the variability in the recovery percentage of the applied fertilizer P by rye plant on different soils. Compared with the maximum buffering capaeity of adsorption (MBCA), MBCD is a better physico chemical index for evaluating the plant-availability of phosphorus in soils. However, it seems that the equilibrium buffering capacity of desorption rnay become the most promising index for diagnosing soil P status.