Interactions of the θ, α, and β EEG rhythms were studied in a set to an emotional facial expression in 30 adult subjects. Electroencephalogram (EEG) data wee analyzed by wavelet transformation in the range 1–35 Hz. The mean and maximum wavelet transformation coefficients (WTC) were computed. Relationships between the β2/α1 and β2/α2 subranges were seen in subjects with erroneous recognition at the set testing stage, as detected using the mean WTC level. The β2/θ rhythm pair showed significant correlation coefficients both for the mean WTC level and the maximum WTC level, and for maximum WTC values in subjects without errors at the set formation stage. New data were obtained supporting the view that the thalamocortical and corticohippocampal systems are involved in supporting the process of recognizing facial expression during set formation and testing, while their interaction may provide a predictor for the success of visual recognition.