Аннотация:The problem of recognizing facial muscles micro-movements, corresponding to various mental states when lying is considered. The objective of the current study is to determine individual mimic indicators of task complexity. The subject of the study was dynamic characteristics of mimic muscles’ micro-movements when answering complex or simple questions. Mimic activity of 12 participants during answering 50 simple and 50 complex questions was studied. The original system of intelligent machine vision let us to digitize respondent’s face muscles micro-movements during the following parts of interview: hearing, thinking, and response (for each question). The R v.3.5.0 language was used for data analysis. Multidimensional statistical methods (factor analysis, cluster analysis, spectral analysis) were applied. Results of 1200 video fragments analysis are presented. High accuracy estimation the subjective complexity of each question for each respondent by facial muscles micro-movement analysis is shown. The implementation of the method into forensic psychological examination practice (or as investigative tool) lets to determine complex for each defendant questions correctly and lead to unerring conclusions.