Our paper analyzes the naturalness of the forest in Babia Hora Nature Reserve, Slovakia, using an objective classification model based on predictive discriminant analysis. The partial indicators of forest naturalness in the model are: the arithmetic mean of the ratio between crown length and tree height, deadwood volume, coverage of grasses, coverage of mosses and lichens, the Clark-Evans aggregation index, and the coefficient of variation of tree diameters. The analysis of the model results revealed that the model was capable of distinguishing the parts of the reserve affected by natural disturbances and/or by past human activities. Hence, the classification model can be used as a supportive decision-making tool when adopting proper measures aimed at improving the actual state or decisions about changing the use of forests with evidently low natural values.