The purpose of the present study was to clarify relationships between logical memory performance on the Wechsler Memory Scale-Revised (WMS-R) (recall of a story presented auditorily immediately (LM-I) and after a delay (LM-II) ) , and other memory functions, including recent memory represented by verbal paired associates (Verbal PA) , and non-memory cognitive functions. The subjects were 44 traumatic brain injury (TBI) patients without dementia or aphasia who were at least 2 months post-injury and underwent the WMS-R, the Wechsler Adult Intelligence Scale Third Edition (WAIS-III) , and several tests of executive function. Scores on the LM-I and LM-II, and Verbal PA-I and Verbal PA-II were correlated with scores on several subtests of the WMS-R, suggesting that each test, whether verbal or visual memory, is interdependent as a memory function. Scores on the LM-I and LM-II were correlated with scores on the Verbal Comprehension Index (VCI) subtests of the WAIS-III, but scores on the Verbal PA-I and Verbal PA-II were not correlated with scores on the VCI subtests of the WAIS-III. Logical memory requires sentence comprehension and narrative interpretation and may depend more on verbal cognitive functions than on verbal paired-associate learning. In multiple regression analysis, the score of LM-I was explained by the score of Picture Completion from the WAIS-III, and the score of LM-II was explained by the scores of Visual Recall-II from the WMS-R and Vocabulary from the WAIS-III. On the other hand, Verbal PA-I was explained by scores on Visual Paired Associates (Visual PA) -I from the WMS-R and Block Design from the WAIS-III, and Verbal PA-II was explained by scores on Visual PA-II from the WMS-R and Letter-Number Sequencing and Block Design from the WAIS-III. No causal relationships were found between the scores of LM-I and Verbal PA-I, or between the scores of LM-II and Verbal PA-II, and some patients showed double dissociations between the scores of LM-I and Verbal PA-I. The patients with impaired verbal paired-associate learning but preserved logical memory tended to show impaired memory functions except logical memory and other cognitive functions. In contrast, patients with “pure” logical memory impairment tended to perform well on most cognitive function tests except for logical memory. Logical memory may be impaired independently of verbal paired-associate learning and other cognitive functions.