Techniques for modelling growth and tree size-class distributions of forest stands have been developed and implemented in an object-oriented program with C++. The object-oriented concept represents high-level abstraction of physical objects in forests and in objects emanating from the data abstraction used for numerical computations of growth. The objects in this forestry context represent relationships of the pines, weeds and site factors and their interactions in a stand, and also use the computational objects. The objects mainly comprise the computations for modelling growth and size-class distribution of tree crops. The percentile interval, a technique derived from the geometry of the interval and its percentile of tree size-class distribution, is used here for modelling tree crops, and the interval with a fixed percentile could be projected over time. This technique could also be used for reconstruction of the distribution to characterize the effects of thinning. The projection of a distribution involves modelling with the initial growth and the long-term changes in tree crops. Initial and later growth could be joined using the algorithm which was derived from an approximation of the linear phase of the cumulative growth function. Experience so far with and future plans for implementation of the developmental work for prediction of forest stands with object-oriented software are also presented.