Two samples of 31 educable mentally handicapped (EMH) children differentiated by IQ level were given a test of verbal problem solving in order to identify the effect of three parameters upon verbal problem-solving performance. The three parameters studied were existential quantification, superordinate set identification, and extraneous information. Eight problem conditions were created, reflecting systematic inclusion or exclusion of each parameter. Six single-step addition problems (addends less than ten) were composed for each of the conditions. Presence of a distractor (extraneous information) was significant (p less than .01). Higher IQ EMH children significantly outperformed lower IQ EMH children on the tasks (p less than .01), but failed to demonstrate expected performance on problems with extraneous information. The findings were felt to suggest a rote computation habit, rather than active processing of verbal information, as the problem solving style of many EMR children at all IQ levels. The need for carefully developed programs in verbal problem solving was indicated. (Author)