The annual girth or diameter growth in the tapping phase is an important trait associated with rubber production, resistance to wind breakage and wood production. The main objective of the present study was to assess the temporal stability of rubber tree genotypes for both natural rubber production and annual girth growth in the post-tapping phase. The phenotypic and genetic correlations of these variables over the years of evaluation were estimated in a rubber tree breeding program. Thirty-two clones were assessed along with the control genotype RRIM 600 for two traits, annual production and girth growth, which were evaluated for five and six years, respectively. A randomized complete block design, with effectively split-plots in time, was used with three replicates, six trees per plot, spaced at 7m×3m. We observed that negative genetic correlations of the accumulated annual girth growth with the accumulated rubber yield (r g =−0.58, P<0.01), and high stability of yield with AMMI statistics explaining 96% of interactions. The study concluded that early selection in the first year of rubber yield may reduce the evaluation time of clones in a rubber tree breeding program. There was a negative phenotypic correlation between annual girth growth and yield. The study allowed differentiation of the genotypes assessed for temporal stability and overall performance for yield during tapping. Genotype selected for stability of production it is not the same as those selected just for annual growth. The stability of annual girth growth correlates negatively with the stability of yield.