BACKGROUND: This study is a replication and extension of a verbal learning experiment reported in this journal (Arkin, Rose, and Hopper, 2000) with individuals who had Alzheimer's disease (AD). Those participants demonstrated implicit and explicit learning of exemplars from an inanimate semantic category (items people wear). AIMS: The purpose of this study was to examine the effects of repeated exposure to animal exemplars on implicit and explicit learning of 13 individuals with mild to moderate dementia due to AD (of which 7 also participated in the earlier study). METHODS AND PROCEDURES: Following 18-20 60-second baseline fluency tests of the target category 'animals', participants engaged in eight sessions of a picture naming and related quiz exercise (study task) that contained 34 words from the target category (exposure words). One hour after each study task session, the fluency test used at baseline was re-administered (experimental fluency test (EFT)). OUTCOMES AND RESULTS: On the study task, the group achieved a significant improvement in the number of correct answers from the first to their best and to their eighth trial (evidence of explicit learning). On the EFTs, 10 of the 13 AD participants produced one or more exposure words never named during the baseline fluency tests (evidence of implicit learning), and 12 out of 13 participants produced novel words that were not exposure words and not produced at baseline (evidence of spreading activation). Additionally, the explicit learning performance of the seven individuals who participated in our earlier study and in this study were compared for the 'items people wear' (nonliving) and 'animals' (living) categories. This comparison revealed a significantly greater improvement in naming performance from the first to the best study task trials for the 'items people wear' (nonliving) category than for the animals (living) category. CONCLUSIONS: Results suggest that repeated exposure to and practice in retrieving category exemplars facilitate short-term (over a 1-hour time interval) implicit and explicit learning and semantic activation in AD participants with very mild to moderately severe dementia. Future directions include exploring the length of time over which this explicit and implicit learning is maintained post-intervention. The differential learning curve for living versus nonliving exemplars observed here is suggestive and requires further investigation under more rigorous conditions.