A new technology era requires new methods. In forestry management the use of stands as convenient compartments of trees are widely spread. These contiguous units of trees of, for example, uniform species composition, height distribution and age are traditionally obtained by delineation of aerial images and further estimations on the characteristics with assistance from field studies. Data acquired from Airborne Laser Scanning systems gives the opportunity to extract each and every visible tree in the canopy. These individual trees are object for automatically grouping into similar clusters as the traditional stands. The aim for this work was to study methods applicable for forest stand delineation. Laser data and Aerial photography as well as results from previously developed analysis methods of Laser data are available in the progress of development of method for stand delineation. This work has resulted in the development and testing of a raster-based clustering method. The data are first processed with an unsupervised ISODATA classifier creating a rough segmentation. Finally, the segments are merged in an agglomerative manner, ie. the best possible pairs of segments are merged first, in all according to intra-cluster variances. The results shows the potential of this clustering method, but in order to create ready-to-use forest stands, more research and development are essential for the whole process.