In 2001 to 2020, for each Long-eared owl nest in the Crane Homeland Nature Reserve territory with an area of 48 km2, the following microhabitat parameters were determined using standard field methods: the nest type built by different bird species, the presence of a nest roof, the nesting tree species, and the nest height above the ground. In most cases, for nesting the Long-eared owl reused magpie nests (Pica pica) (77%), and half of the nests were found on willow trees (family Salicaceae) (50%). The nest height above the ground varied from 1.5 to 22 m. Vegetation on the nesting area with a radius of 500 m was determined using the “Semi-Automatic Classification Plugin” automatic vegetation classifier of QGIS software. Five habitat types were so classified: forests, shrubs, meadows, arable land, and satellite image mistakes. Village areas were allocated separately. The average shares of the largest habitat types for over 20 years on the nesting sites included meadows (51.1%) and shrubs (16%). From 2001 to 2020, a habitat change in the breeding area was observed: forest and shrub areas decreased, while the meadow, village and arable land areas increased. An analysis of the effect of habitat types on the nesting success was evaluated by means of the “gbm” and “dismo” modules in the R program using the Boosted Tree classifier algorithm. The presence of such habitat types in the nesting area as villages (24.2%), shrubs (19.3%), meadows (11.8%), and habitat heterogeneity (11.6%) most strongly affected the probability of owl successful nesting. That owls tend to use anthropogenic and agricultural territories is associated with the diversity, easy foraging and more effective protection from predators during the breeding season.