This experiment was carried out in order to know the decreasing causes of plant density and its relations with pasture productivity by the times of established pasture. The results were summarized as follows, 1. The yield components of dead plant can be used to estimate the changes of pasture productivity by the times of established pasture. 2 . The number of dead plants(PD) showed a quadraticalrelation with mean dry weight of dead plant $W/pl.)$ by the times of established pasture. 3. Relationship between relative number of tillers per dead plant to mean number of tillers of adjacent plants (RNT) and relative ratio of number of tillers per dead plant to number of tillers of plants in the 1st cutting (RNDT) showed highly positive significant correlation in 1 and 3 years old pasture. 4. Less than 100 of RNT were 79.5 %, 71.4 % and 100 % of dead plants in 1, 3 and 5, and 7 years old pasture, respectively. 5. Not exceeding values of RNDT to MRRNT were 89.7 %, 71.4 % and 100 % of dead plants in 1, 3 and 5, and 7 years old pasture. 6 . The values of 79.5 %, 57.2 % and 100 % of dead plants in 1, 3 and 5, and 7 years old pasture were not exceeding to MRRNT and less than 100 of RNT. 7. From the results described above, it was discussed that the relative number of tillers per dead plant(RNT) should be described not absolute deaeasing causes of plant density in 1 and 3 yars old pasture. But, the total number of dead plants in 5 and 7 years old pasture were absolutely influenced by RNT and MRRNT.