The article presents the results of experimental studies aimed at testing the general hypothesis: when using comparative judgments, there are differences in the effectiveness of the organization of mental actions and the subsequent reproduction of their semantic content, depending on the type of comparative relations being established, and also on the features of the syntactic structure of verbal instructions. As independent variables in the experiments used problems that were formulated with the help of comparative judgments. The content of the tasks included different types of comparative relationships. At the same time, with each kind of comparative relationship, different types of logical-syntactic organization of comparative judgments were used. The number of correctly solved problems was recorded depending on the compared spatio-temporal and subject-semantic relations, as well as on the types of syntactic structures used. The obtained results confirm the general hypothesis. The most effective in the mental solution of problems are established “faster – slower” and “higher – lower”, and less effective “left – right” and “near – farther”. At the same time, the most successfully solved problems that require the mental establishment of comparative relationships in the case when the image of the object, the choice of which serves as a solution, is constructed by the logical-syntactic organization of verbal instructions is constructed by the last of the three compared objects. The least successful are the tasks that are characterized by such a logical-syntactic organization, in which the orientation in establishing comparative relationships presupposes one of two options for the mental rearrangement of the represented objects in a linear sequence (C > A > B or B < A < C) as a result of which the location of the object In the represented series can be the first or the last. At the same time, the results obtained indicate that there is no age-related dynamics in the spontaneous development of the skills to establish different types of space-time and subject-semantic relationships in mental activities in the age range from 12 years (schoolchildren of the 5th grade) to 21–22 years (students 4-th courses). Also, there were no gender differences in the success of solving different types of tasks.