One of the most significant advantages of intelligent tutoring systems is their ability to adapt to the current knowledge level of each individual learner by offering tasks that are most suitable for him/her at the current phase of learning process. The main problem is how to evaluate the degree of task difficulty, which, as a rule, is done subjectively. The paper presents results of ongoing research on evaluation of concept map complexity. The latter is necessary for more discriminative estimation of the degree of task difficulty and development of more accurate scoring system used in already developed intelligent concept map-based knowledge assessment system IKAS. The basic idea is to interpret for concept maps and to use for evaluation of their complexity the four aspects applied for estimation of systems complexity – the number of systems elements and relationships between them, attributes of systems, their elements, and relationships, and the organizational degree of systems. A method for determination of structural importance of concepts in a concept map is also proposed. The approach is described using examples of relatively simple abstract hierarchical concept maps represented by incoming trees.