Latex Diagnosis (LD) is currently considered by Cirad and most of its rubber agro-industry partners as a routine physiological tool to optimise, at block level, the rubber yield production of the rubber plantations. Without using LD, a general stimulation recommendation per clone and per tapping year is generally applied at plantation scale, as a function of tapping cut position and direction, whatever the local and actual yield potential is. Even though this general recommendation is based on clonal physiological latex characteristics, such a global approach does not permit to consider the local specificities of the yield potential, as it avoids considering factors like soil heterogeneity, microclimate variations in larger estates and differential expression of diseases (leaf diseases, root diseases...). In this case, plantations are almost "blind" regarding suitability of the applied stimulation intensities, and uniform application of the same rate of stimulant in all homogenous cultural units may sometimes lead to optimised exploitation but may also lead locally to underexploitation in higher yield potential areas or to overexploitation in lower yield potential areas. Using LD permits to optimize the stimulation at local level (decrease of stimulation when an overexploitation is detected, increase of stimulation intensity when an underexploitation is detected) and therefore permits the yield optimisation block per block, taking into account the plantations heterogeneities and therefore the actual local yield potential. Of course, LD interpretation depends on former set up LD parameters reference values. These ones are clonal and established for the 4 parameters used in LD: latex sucrose content, latex inorganic phosphorus content, latex reduced thiols content and DRC/TSC. These LD reference values are established for 5 limit levels (very low, low, normal, high and very high), for each LD parameter (Suc, Pi, RSH and DRC/TSC), either at regional scale or, in case o