This study was undertaken to investigate whether alcoholics differ qualitatively from controls in their organizational approach to the Rey–Osterrieth Complex Figure (ROCF) and to examine the effects of problem-solving and organizational ability on learning and retention. Data were collected on 3 groups of neuromedically healthy, middle-aged men: 29 recently detoxified alcoholics (RDA) abstinent at least 2 weeks, 29 long-term abstinent alcoholics (LTA) abstinent at least 18 months, and 29 nonalcoholic controls (NAC). Groups were comparable in age, education, WAIS–R Vocabulary score, and ANART IQ. Three indices of problem-solving approach were used to score how participants produce the ROCF during the copy condition: perceptual clustering, organization, and constructional accuracy. Learning and retention were measured by immediate recall, 20-min delayed recall, and recognition performance. RDAs had significantly impaired immediate recall, delayed recall, and recognition scores compared to NACs. RDAs also had significantly lower perceptual clustering, organization, and constructional accuracy scores than their long-term abstinent and nonalcoholic counterparts. Multiple regression analyses revealed that problem-solving ability at copy accounted for a significant proportion of the variance in immediate recall, delayed recall, and recognition. Further investigation of memory performance using a repeated measures design revealed that across the 3 groups, little forgetting occurred between immediate and delayed recall, while performance improved significantly on the recognition trial. Use of less efficient problem-solving strategies appears to affect RDAs' ability to learn and retrieve complex figural information. (JINS, 2000, 6, 12–19.)