An experiment was designed totest the hypothesis that suppressionof the amplitudeof thealpha rhythm provides a reliable objective index of visual imageryunder two conditions of problem solution, namely, solving a 'verbal' problem and solving a 'visual' problem. Sixty-nine subjects were examined in the following way: first, preliminary 'runs' were made with each subject for EEG recording under the It was found that, although the visual content of awareness during the two problem-solution periods as reported by the subjects differed markedly, suppression of alpha rhythms occurred under both conditions, the amplitudes in both caaes falling about half-way between the two values recorded in the preliminary 'runs'. There was, however, a greater suppression effect in the caae of the 'visual' problem. The results, aa a whole, therefore, suggest that the hypothesis is not tenable, since 'imagery' appears to be only one of many factors that may produce suppression effects. This conclusion is supported by a breakdown of the subject sample, according to reports of imagery on the verbal problem, which showed that there was no difference in the basic patterns aa between those who visualized the verbal problem and those who did not.Repeated solution in a check period, immediately following the solution period, made no significant difference to suppression of alpha in the verbal problem, but did so in the visual problem. The suggested interpretation follows Fisher (1916) in regarding this as a possible decrease in image content.