Physical and mental disorders in people affected by the consequences of the Chernobyl accident have worsened rather than declined with time [7, 9, 12]. Central nervous system diseases accompanied by impairment of higher mental functions are among the leading disorders in this sphere. Comparative analysis of changes in electroencephalograms (EEGs) and neuropsychological characteristics of the participants of the clean-up of the Chernobyl accident consequences demonstrated that, at the initial stages of pathology formation, predominant damage was inflicted to subcortical structures of the limbic‐reticular complex, whereas the cerebral cortex was involved in the pathological process during further stages of its development [9, 12, 15]. In addition to a definite similarity between neuropsychological syndromes observed in the participants of the clean-up of the Chernobyl accident consequences, there were features of substantial difference between these syndromes determined by the level of the location of dominant brain lesions, which could be observed, e.g., in EEG patterns [12]. The goal of this work was to study the quality of solution of certain cognitive tests by the Chernobyl accident clean-up participants with different types of EEG. Electroencephalograms were recorded in 24 participants of the clean-up of the Chernobyl accident consequences (men; dextrals; mean age, 44.5 ∠ 3 years) at rest and during subsequent performance of three cognitive tests. A group of 20 apparently healthy subjects (men; dextrals; mean age, 40 ∠ 6 years) was used as a control. EEGs were recorded in 16 or 18 leads (with a common ear electrode) for 60 s at a sampling frequency of 100 Hz. Power spectra of EEGs were calculated within the frequency band from 0.5 to 40 Hz for six isolated ranges of EEG rhythms (delta, theta-1, theta-2, alpha, beta-1, and beta-2) and 10‐15 epochs (2.56 s each). The BrainLoc program (MBN, Russia) was used to analyze the three-dimensional localization of the dipole sources of paroxysmal bioelectrical activity. This program was described in more detail in [1‐3, 15]. The following computer-assisted procedures were used to assess the cognitive capacity: Counting at a Given Rate, Raven’s Progressive Matrices, and Tracing tests. The test tasks were presented to healthy subjects and patients on a monitor screen [11]. The Counting at a Given Rate test enabled researchers to examine simple logical activity and short-term memory. The subjects were instructed to perform simple mental calculation with numbers from 0 to 9 randomly presented for 1 s on the monitor screen. Green and red numbers should be added and subtracted, respectively. Raven’s Progressive Matrices test (the monochromic variant) was used to assess spatial‐visual thinking. The subjects were instructed to find an analogy between two pairs of