A classification of spectral patterns of EEG underlies several cognitive neurotechnologies including passive and active brain-com puter interfaces. Despite arithm etic tasks often being used in studies of cognitive workload, there is a lack of findings describing a possibility to recognize EEG patterns related to different types of m ath operations. In th e present work, we have shown th a t th e power spectral density of EEG can be used to classify types of m ental operations including a classification of verbal and different m athem atical tasks for simple arithm etic operations or logical tasks w ith arithm etic progressions. The verbal tasks were separated from arithm etic ones significantly b etter th an arithm etic from logical tasks, and verbal from logical tasks. B etter discrim ination of verbal tasks from arithm etic b u t n ot from logical tasks supports the hypothesis of unique EEG patterns associated w ith verbal activity th a t apparently differ from m ental operations in arithm etic. Additionally, we compared th e behavioral performance in problem solving and accuracy of EEG classification in tw o groups of subjects w ith education in m ath or hum anities (N = 8 + 8). We obtained th e predicted differences related to b etter performance of th e m ath group in solving m ath tasks th an th e hum anitarian group. However, the classification accuracy of tasks based on EEG did n ot differ significantly between groups and was essentially higher th an random. Considered together, our results support the hypothesis th a t EEG patterns reflect individual cognitive states corresponding to m ental operations and can be used in classification of different cognitive activity.