At the 3 sites visited, the leaf sampling method was set in place, from sampling in the field to dispatching, and including conditioning, drying and packaging. The 3 plantations are currently being structured into Leaf Sampling Units (LSU). It is on that structure that the fertilization policy is based. Cross-checking of the 26 leaf samples analysed in duplicate by CIRAD showed that the results provided by the local laboratory (Bah Lias Research Station) varied depending on the nutrient analysed, but were generally underestimated, and in particular the coefficients of correlation were not very good. For each nutrient, a linear formula can be used to calculate a corrected value from the raw value. Generally speaking, the nutritional status of the palms is good in 2006 at the 3 sites, and fertilizer application rates can be reduced in 2007 compared to those applied so far, without taking any risks on yield potential. A fertilizer schedule is proposed for 2007. For each nutrient, its corrected value is first calculated, then that value is entered in the leaf content-fertilizer rate correspondence table, thereby giving the quantity of fertilizer to be applied per palm on all the palms in the same LSU. Once the rates per palm have been determined, the quantities of each type of fertilizer are summed up per LSU then per division and per site, which gives the respective tonnages to be ordered for the 2007 season. A reminder is given of a few fertilization rules. Fertilization should be finished by the end of June 2007 at the latest, so that leaf samples can be taken at the beginning of August 2007. The locations for the future reference mineral nutrition trials have been chosen at each of the 3 sites. Soil samples were taken at Agrowiyana and Kisaran, for which the analysis results are commented on. A soil sample will be taken in February 2007 at Pasaman. In principle, it will be possible to launch 3 trials in 2007.