California Verbal Learning Test (CVLT) performance was compared for 33 head-injured and 33 neurologically normal adult males. On learning trials, head-injured persons recalled fewer items than controls in general, though the percent of improvement between trials did not differ for these groups. Head-injured patients also demonstrated more intrusions and used semantic groupings less than did normal controls during learning trials. On delayed recall trials, head-injured persons remembered a smaller percentage of items they had learned earlier than did controls. Semantic cues aided recall for head-injured but not normal persons. Head-injured persons were less able to discriminate list from nonlist items during the recognition trial. Results indicated both retrieval and storage (i.e., encoding or retention) deficits for the head-injured group. The CVLT is a useful instrument for defining memory deficits in a head-injured population.