This study presents a method to spatially predict stand composition on tree genus level using logistic regression models based on both analogue and digital airborne remote sensing data and field samples. The prediction is based on 20 explanatory variables derived from spectral information of ADS40 images, CIR RC30 aerial images, high-quality DSMs derived from it and LiDAR DTM and DSMs. In a first step, canopy height models (CHM) were generated by adjusting and subtracting the DSMs from medium footprint LiDAR DTM. In a second step, fractional shrub/tree covers were modeled using logistic regression and explanatory variables derived from the digital surface models. In a third step, stand composition was modeled by extracting groups of trees and single trees using a combination of image smoothing and image segmentation. Then, for each pixel of a segmented tree or tree group probability belonging to one of the eight main tree species was predicted (Picea abies, Larix, Pinus sylvestris, Abies alba, and Fagus, Fraxinus, Betula, Acer). For this step, a second logistic regression model was used with additional spectral information obtained from the channels of the scanned aerial images and the digital sensor images as explanatory variables. Good results were obtained for the extraction of forest area and distinction between deciduous and coniferous trees (kappa of > 0.8). Prediction of main tree species was validated using 480 tree objects from field surveys and revealed satisfactory results (kappa = 0.5 to 0.8). Poorest agreement was obtained for deciduous trees. This study underlines that high-quality DSMs and spectral information from the RC30 CIR aerial images are sufficient for prediction of tree genera on forest stand composition level. However, this study reveals also that reliable prediction of main tree species can only be obtained by the additionally usage of airborne digital sensor data. The obtained information on stand composition and main tree species is helpful for assessing the impact of certain tree species on the non-forest areas in a mire biotope as needed by a sustainable mire habitat management.