If, as suggested, creative (insight) problem solving is less systematic and employs less planning than analytical problem solving, the former requires substantially less working memory (WM) than the latter. Subjects simultaneously solved problems and counted auditory stimuli (concurrent WM task), in response to which ERPs were recorded. Counting disrupted analytical, but not creative performance. Peak and time-window average P300 were more frontal during analytical problem solving as compared to insight or counting tones only (control). A PCA extracted two factors in the P3 range, one frontal and one broad left-lateralized, which distinguished analytical from creative problem solving. The findings indicate distinct processing pathways for the two types of tasks with more WM involvement in analytical tasks.