The study presents the changes of structure in three Norway spruce-dominated forest stands with different soil type and aspect in demonstration object Pro Silva Donovaly-Mistríky (the Low Tatra Mts., Slovakia). In surveyed stands, the conversion to single-tree selection system using the current generation is being applied for the period of 40 years. Based on the data from the full calipering in 1979 and the stand inventory (stems and natural regeneration) in 1999, 2013 and 2021 on the series of permanent research plots (0.25 ha) in investigated stands, a relatively rapid advancement in the conversion process was observed. An appropriate model of selection forest is characterized by the optimal stand volume 320 m³ ha-1 and target dbh 66 cm in protection forests on ranker soil and by stand volume 385 m3 ha-1 and target dbh 66 cm in stand with dominant production function. Under the ongoing climate change, the high proportion of Norway spruce in the stands on nutrient-poor soils and the southern aspect could lead to their fast disintegration and gradual loss of differentiated structure. Research objects of this study were three forest stands with dominant proportion of Norway spruce in demonstration object Pro Silva DonovalyMistríky in orographic unit Nízke Tatry (Slovakia). Forest stands with different soil type and aspect have been converted to selection forest using the current generation for more than 40 years. The results of full calipering in 1979 and the stand inventory in 1999, 2013 and 2021 on the permanent research plots (0.25 ha) confirm that the stands have reached the stage of structuration (compartment 1606 and 1632c) or the stage of structure achievement (compartment 1631). In the compartment 1606 with southern aspect, despite the enhanced conditions of ranker soil, we observed a rapid dieback of spruce in the upper tree layer of the converted stand. In this phase, the stability of the upper tree layer is guaranteed by silver fir, as its vitality in these ecological conditions is acceptable. During the investigated period, dynamics of regeneration processes (individuals in height range 50–130 cm) was adequate, with balanced representation of spruce (58%) and fir (42% – Tab. 3). On the other hand, a significant decrease of tree density was registered in the category of ingrowth (height ˃ 130 cm, dbh ≤ 8 cm), with the reduction from 4480 ha-1 in 2013 to 1560 ha-1 in 2021. Despite the decrease of fir proportion (77%) in favour of spruce (23%), the dominance of fir in the lower tree layer is maintained (Tab. 3). In compartment 1632c, the ranker soil played a significant role in differentiation of diameter structure and natural regeneration of spruce and fir in clusters (Fig. 1 and 2; Tab. 3). The dynamics of regeneration processes during the investigated period was continuous, with the spruce proportion in natural regeneration (height 50–130 cm) reaching 53% and the fir proportion of 38% in 2021. The decrease to the density 1720 ha-1 was observed in the category of ingrowth. Reduction of the trees in this category that provides the individuals for the diameter class 10 cm resulted from the high tree density in the diameter class 10 cm. In compartment 1631, the dynamics of regeneration processes during the surveyed period was adequate, with balanced proportion of spruce and fir. The significant decrease (by 200%) in the ingrowth category was registered also in this compartment (Tab. 3). In the investigated stands, the reduction of the annual volume increment with the gradual approaching to the selection structure was observed. On the permanent research plots in compartments 1606 and 1632c that represent the stage of structuration, the annual volume increment reached 4.5 m³ ha-1 and 4 m³ ha-1. Compartment 1631 that is in the stage of structure achievement has the annual volume increment of 7 m3 ha-1. According to the ongoing continuity of regeneration processes and the differentiation of stand structure, the model with optimal stand volume of 320 m³ ha-1 and target dbh of 66 cm proposed for the protection stands (compartments 1606 and 1632c) was confirmed as appropriate. Model with such parameters is able to maintain the sustainable fulfilment of the non-production functions as well as the high-quality timber production (resonance wood). In the stand with dominant production function (compartment 1631) it is possible to set the target diameter to 66 cm and the optimal stand volume to 385 m³ ha-1 (Tab. 3; Fig. 1). The results confirmed that on such high-productive site even in the case of significant exceeding of the optimal stand volume this does not represent the risk for the regeneration processes. [ABSTRACT FROM AUTHOR]