Studies of localized brain dysfunctions have revealed connections between patterns of cognitive dysfunction and specific profiles of memory impairment. The amnesic type of memory impairment is defined by deficits in both free recall and recognition memory, whereas the dysexecutive memory impairment is characterized by retrieval deficits, i.e. a disproportional impairment in free recall relative to recognition memory. The present study tests whether classifications of psychiatric patients into recall impaired only (= RO group) and Recall and Recognition impaired (= RR group) correspond to the executive type and amnesic type of memory impairment. The alternative hypothesis is that the two groups merely differ in degree of neuropsychological and psychiatric disturbance. Forty-four subjects impaired on California Verbal Learning Test (CVLT) were selected from a larger database of 103 impaired and non-impaired subjects with schizophrenia or recurrent major depression. Subjects were classified into RO and RR groups and compared on measures of memory strategy (recency effect and interference on CVLT), overall neuropsychological function (Stroop Test and WAIS-R similarity) and psychiatric symptom load (positive and negative symptoms). Repeated measures ANOVA showed no effects of group, i.e. the RR group did not perform consistently below the RO group with regard to memory strategy, neuropsychological function or psychiatric symptom load. Two out of three analyses showed group x test interaction, supporting the dissociation of distinct executive and amnesic profiles among psychiatric patients. The RO group was more susceptible to interference but had better recency score than the RR group. The RO had higher negative symptoms while the RR group had higher positive symptoms.