The aim of the paper was to examine the current possibilities of digital aerotriangulation of the aerial images taken by the large format (Ultracam X - project 1, Xp - project 2) and medium format (Leica RCD30 - project 3) digital cameras, when using GSD 0.10 m (large format cameras) and GSD 0.20 m (medium format camera) used in the forestry mapping practice. To verify the accuracy of the aerotriangulation, terrestrially measured checkpoints were used. In general, the best results of the horizontal and vertical accuracy were achieved in solutions, where more control points were used and GPS/IMU corrections were applied. Also the impact of camera self-calibration parameters to correct systematic shifts in the photos was proven. The best achieved overall results characterized by the root mean square error (m(xy) and m(z)), were m(xy) = 0.047 m, m(z) = 0.109 m in project 1 and m(xy) = 0,058 m, m(z) = 0.159 m in project 2. When comparing results of aerotriangulation with 4 control points and the best achieved solution in project 1, the difference was minimal (project 1 m(xy) = 0.051 m, m(z) = 0.121 m and in the project 2 m(xy) = 0.064 m, m(z) = 0.165 m). The significant impact of the absence of control points in the solution nr. 4 in each project was mainly in the vertical accuracy (m(z) = 0.266 m - project 1 and m(z) = 1.052 m - project 2). The best achieved result in the project 3 (Leica RCD30) used 19 control and 19 check points with m(xy) = 0.21 m, which is at the spatial resolution of used aerial photos. Achieved accuracy fully complies with forestry mapping in Slovakia, where expected accuracy of the determination of the elements of forest mapping is up to 0.5 m.
