Difficulty of picture and word recognition was manipulated to test the hypothesis that schizophrenics have adequate recognition memories and to determine whether overinclusion is related to a good visual recognition memory. Good and poor premorbid, acute and chronic schizophrenics were compared to non sychotic psychiatric atients and to hospital aides. Both Payne's Object Rassification Test a n t Goldstein-Scheerer's Object Sorting Test were given to all the patients, but only the former differentiated among the diagnostic subgroups and was related to good picture memory. When the measures of overinclusion, chronicity, premorbidity and intelligence were used as independent variables in a multiple regression, only Payne NonA predicted success on the picture task. Good premorbid schizophrenics recognized pictures as well as normals and nonpsychotics, but poor premorbids' picture memory was significantly worse. No experimental variable predicted word performance in a multiple regression, and good premorbids did not differ from poor premorbids on their word recognition. The authors favored the interpretation that the pattern of performance of overinclusive schizophrenics reflects their tendency to scan broadly both important and unimportant features of stimuli.