To find out the residual effect of phosphate fertilization on botanical composition, mineral contents and mineral nutrient deprivation of pasture plants, this experiment was arranged as a randomized complete block design with six treatments(0-0, 50-50, 100-65, 150-65, 200-65 and 250-65kg $P_2O_5$/ha), those were composed of three P,O, fertilization level(0, 50 and 65kg P,OJha) after phosphate fertilization trial with six treatments(0, 50, 100, 150, 200 and 250kg $P_2O_5$/ha) from 1989 to 1992, and conducted at hilly land in Kwangju, Kyonggi Province 60m 1993 to 1994. The results obtained are summarized as follows: The botanical composition of pasture plants without phosphate fertilization was very poor. With increasing available phosphate of soil, the percentage of grasses and legumes were remarkably increased and weeds was decreased. With 150-65kg $P_2O_5$/ha application for two years, the average DM yield was composed of 80.6% grasses, 15.4% legume and 4.0% weeds, the percentage was similar to that of 200-65kg $P_2O_5$/ha. As available phosphate of soil increase, P, K and Mg content of pasture plants were increased, but Ca/P ratio was lowed and the other mineral contents tended not to be regular. The mineral nutrient contents deprived by DM yield was increased as available phosphate of soil was increased, but those were not different between 150-65 and 250-65kg $P_2O_5$/ha. However, mineral nutrient of $P_2O_5$,$K_2O$ and MgO were higher in 200-65 and 250-65kg $P_2O_5$/ha. Most of applied mineral elements were accumulated at a subsurface(0-2.5cm) of grasslands, and mineral tended to be decrease than that of soil chemical analysis before the experiment of residual effect of phosphate fertilization.