Within this project three different objectives have been the target ofinvestigations. 1. Species grouping. Based on expert knowledge a groupingof 436 tree species occurring in Sabah rain forests was undertaken in 5height groups (criteria 1) and 3 light demand groups (criteria 2). Anindependent performance of both groupings results in max. 15 final groups,from which three (lowest height layer) are considered as identical. Endingup with the resulting 13 groups this new species grouping should be the basisfor a further development of the FORMIX4 model. Height-to-diameter-curvesstill have to be improved for this new grouping. 2. Regeneration: Data frominventories in four different forest reserves in Sabah were analyzed forregeneration of tree species. Because the field data did not content thetree size we are interested in (h <= 1.3m) we can only end up with someorders of magnitude and general features about regeneration. Additionallyinteresting publication and Ph.D. thesis were analyzed. As a result itseems that the number of seedlings varies widely and depends on the distanceto the mother trees, the number of mother trees and a biological phenomenoncalled mass fruiting which occurs every 4-6 years. Some functionalrelationships between number of seedlings and their occurrence in theFORMIX4 model can be mentioned, however the defined numbers of seedlingsare better taken out of the research undertaken parallel to this study byMr. Klaus Werner. His results combined with the expected range for smalltrees might give a good basis for site quality dependent regeneration pattern.3. Mortality: Data from permanent sample plots in four different forestreserves over different lengths of time were analyzed for relationshipsbetween annual mortality rate and functional group, tree size, tree growthand various combinations of the latter. Some of the data sets were stronglyinfluenced by a drought in 1982/83 which makes it impossible to presentreliable figures for the mortality. Therefore again only trends can beworked out. Mortality seems to be significant higher in pioneer species,for trees with no or negative increment and seedlings depending on theirtree density. All other relations are not confirmed over the whole rangeof the data set, but might be used in single case studies. They are:Decreasing mortality rate in big trees, especially for single functionalgroups (pioneer and emergent). Decreasing mortality rate in trees withlarge annual diameter increment. Increasing mortality over a short perioddue to environmental circumstances (e. g. droughts, El Nino ). As generalit can be said that there are very few principle pattern in regenerationand mortality which can be applied for simulation of rain forest. Aworthwhile investigation might be the analysis of a data set of a hugearea over a long period of time like the established 50ha plots in Lambir,Sarawak and Pasoh, Peninsular Malaysia. Unfortunately these data set werenot available for this study.