The use of three-dimensional geovisualization is currently pushed both by the technology development and users’ demands in different areas of human activity. However, the usability and users issues of 3D are still ambiguous. In recent years, several research groups analysed the empirical studies on 3D visualizations from different viewpoint including the conceptual aspects influencing the results (Coltekin et al 2016), different types of immersion and interactivity (Špriňarová et al 2015), and also the role different interaction with the 3D environment (Herman et al 2016). Some of these papers were exploratory and revealed the potential problems to be addressed as the necessity to conduct further empirical studies in order to understand what specifically the usefulness and usability of 3D visualizations depends on. We decided to further develop the previous study (Špriňarová et al 2015). Research questions were defined as follows: • Can the level of immersion influence particular user when using pseudo 3D versus real 3D visualization? • How the interactivity influences the perception of 3D visualization? • What is the role of personal spatial abilities in the process of solving 3D visualization tasks? We used the within-subjects experimental design, but the probands were divided into two groups which worked on informationally equivalent tasks in different order. Probands consisted of the group of 20 high school students aged 16 – 17 (12 M/ 8F). The experiment consisted of series of tasks measuring the ability of participants to identify the correct terrain profile in three-dimensional digital elevation model (DEM) visualization. Probands were asked to identify the correct profile (one of the three) which correspond to the line between two points on the DEM (task type 1:3) or identify which of the three lines displayed on DEM correspond to the presented profile (task type 3:1). There were static and interactive parts and tasks were visualized in pseudo and real 3D mode during the experiment. The test was supplemented with a psychological test of spatial abilities (mental rotation test) and think aloud protocol afterward. Speed and correctness of the answers were recorded. Preliminary results support the fact, that the usability of real 3D technology for the interactive VGE remains ambiguous and is task dependant because the probands performed differently in 3:1 and 1:3 tasks. It also shows that tasks performed on the real 3D visualization are the slower in average contrary to the tasks performed on the pseudo 3D visualization. Further research will be focused on the influence of gaming and extension of probands group with the people from different age categories.