Yield adaptation was assessed in twelve rubber clones evolved in India (RRII 105, RRII 118, and RRII 208), Malaysia (RRIM 600, PB5/51, PB 86, GI 1) Indonesia (PR 107, GT 1) and China (SCATC 88/13, SCATC 93/114, Haiken 1). The yield data of sub optimal and optimal environments were seperated into four enviroments of two successive years (1999-2001). GE biplot was made to assess adaptability of clones to the environments. The largest positive value was seen for RRII 208, indicating RRII 208 performs well across environments. All the other extreme is RRII 105 that extended to the negative side of vector A. When all clones are considered together, the order of interaction from most positive to most negative are RRII 208, SCATC 93/114, PB 86, RRIM 600, PR 107, RRII 118, GT 1, PB 5.51, HAIKEN 1, SCATC 88/13, GI 1 and RRII 105. The genotypes are well spread through the diagram and hence no obvious grouping of the clones is evident. To excercise selection based on both stability and yield, the PCA 1 values were plotted against individual genotype means. PB 86 had a genotypic mean which was greater than the general mean (31.49 g tree -1, tap -1) and the PCA value is almost zero, thereby indicating that PB 86 is stable over all environments. The clones exhibited same trend of stability when PCA 1 values were depicted against environment means. RRII 118, RRII 208, RRIM 600, PB 5/51, SCATC 88/13 and SCATC 93/114 were seen to interact positively with varied environments since their position were above zero value when depicted against both genotype and environment means.