Ten two hectare forest plots were established on two distinct reserves, one mainly on volcanic deposits and the other on sedimentary and alluvial soils. These were investigated for their site carrying capacity and species diversity in relation to edaphic factors, using accumulated basal area as the growth indicator. A soil survey was conducted in each reserve to classify the soil types at the series level. Five two hectare sites each reflecting different soil types were chosen from each reserve. All trees of 10 cm dbh and over were enumerated on each soil type and identified to species level. Fourteen tree species were selected for foliage sampling. A representative soil pit was dug for each soil type and the different horizons were sampled. In addition, ten composite samples were randomly selected to represent soil depths 0-15 cm and 15-30 cm. A complete physical and chemical analysis was carried out on the samples. The results clearly exhibited the influence of parent material on particle size distribution, soil colour, bulk density, porosity, moisture retention and soil chemical compositions. In addition the phosphate fixation problem was discussed. Soil fertility is governed by the parent material from which the soil is derived and is parallelled by the site carrying capacity of the tropical rain forest. Reserve of K was demonstrated to be the main growth limiting nutrient. This was well supported by both soil and foliage analyses data. The nutrients N and P also influence growth but in these cases consistent relationships with basal area were only established after removing some out-lying points. The available and exchangeable soil nutrients were found to be poor indicators of growth. There appears a possibility of using Shorea leprosula as indicator species for foliar analysis in order to assess site fertility. Certain physical properties such as bulk density, clay content, site gradient and available water also exerted some influence on the growth of the trees. Fertile sites contained less species than nutritionally poorer sites. The competition-domination-suppression phenomenon is put forward in explanation. The geological body from which soil is derived is more dominant in determining species diversity than is the soil series per se. On poor sites dipterocarp species are more diverse and they also grow better on lower gradients. Chance factor plays more of an indirect role in species diversity and is believed to be very critical (especially on fertile sites) in the regeneration of dipterocarp species.