Despite short-term fluctuation of rubber prices, the demand of natural rubber has increased continuouslywith the increase in population and living standards of the human being. Nevertheless, urbanizationresults in the fast depletion of forests as well as the land area under rubber. Therefore, in order tomeet the continuous demand on latex, the productivity of rubber plantations should be increased.While producing high yielding clones for improved yield per tree which is a long-term process inperennial crops, planting density could be adjusted to obtain high productivity in rubber plantations.The present level of planting density of rubber in Sri Lanka has been decided on the experimentsconducted with the genotypes which are not in common use at the moment. Also, the optimum densityshould vary with different socio-economic conditions. Therefore, the present study was aimed toidentify the suitable planting density for the recently developed and commonly used genotypes of rubber. Specifically, this paper is focused to assess the overall productivity of rubber with respect tolatex yield.The experiment was set up in Rathnapura district of Sri Lanka in 1992. Rubber was planted in threehigh densities, i.e., 600, 700, 800 trees per hectare, with the presently recommended level of500 treesper hectare. Also, 3 genotypes (clones) i.e., RRIC 100, RRIC 110 and RRIC 121 were incorporatedwith the statistical design of split plot where the planting densities were laid as the main plots whilstclones were in the sub plots. Growth and yield parameters in terms of girth, bark thickness (8T), andthe incidences of tapping panel dryness (TPD) were assessed yearly and the latex volume (LV), %dry rubber content (DRC), number of trees in tapping (TIT) were assessed on daily basis up to 2004.Irrespective of the clone used, mean yield per tree per tapping decreased with increase in plantingdensity. However, it was vice versa in the case of yield per hectare (YPH) due to the increase in TITwith increase in planting density. The percentage of trees with 'fPD was not significantly affected bythe planting density. The clone RRIC 110 was infected with Coryncspora leaf disease hence poorperformance was shown in all densities. The trend of increasing YPH was similar in both otherclones. i.c. RRIC 100 and RRIC 121. However, the rate of increase in TIT showed a decline of700trees per hectare resulting in a lower YPII at this level than expected. Although latex productivitycould generally be increased with increase in planting density, overall economic profitability of thesystem wi II also depend on cost of production, amount and va lue or tim bel' and carbon produced.Therefore, study warrants further investigations on above issues before making any changes to thepresently recommended planting density.