Estimation of the production and dynamics of tropical rain forest is an important issue for a sustainable management of these forests. In this paper we use the forest model FORMIND for analysing logging strategies. The model describes growth, mortality, recruitment of trees and competition between trees. The calculation of tree growth is based on a carbon balance. Dying large trees fall down and create gaps in the forest. Tree species are grouped in plant functional types. Different m anagement strategies for Dipterocarp, rain forest in Malaysia were analysed. The scenarios differ regarding the logging cycle, logging method, cutting limit and logging intensities. We characterise the impacts with three criteria describing the yield, canopy opening and changes in species composition. Multicriteria decision analysis was used to evaluate the scenarios and identify the efficient ones. Our results plainly show that reduced-impact logging sc enarios are more 'efficient' than the others. Nevertheless there is a trade-off between yield and achieving a desired ecological state of logged forest; the ecological state of the logged forests can only be improved by reducing yields and enlarging the logging cycles. Our study also demonstrates that high cutting limits or low logging intensities cannot compensate for the high damage caused by conventional logging techniques.