Several tests were carried out of unmanned aerial vehicles (UAV) equipped with customer cameras to take pictures of forest areas. Based on this a special own UAV was built which can carry a set of non-metric digital cameras recording images in RGB and near-IR channels. Approximately 1000 aerial images with a ground resolution of 0.15 m were taken for the selected research facility with an area of approximately 250 ha. These images were used to produce an orthomosaic in natural and multispectral bands. It was decided to find out whether it is possible to evaluate the health condition of spruce (Picea abies). The field observations were made of the crowns of spruces growing in about 40-year-old single-species stand. Information on the density of pine needles, its color, and other symptoms of the tree condition were recorded. The same trees were identified on orthomosaic and classified. Results of classification from the field and orthomosaic were compared. It was concluded, that it is possible to divide the observed number of trees into several health condition classes.