This study examined the validity of semantic clustering indexes developed for the Hopkins Verbal Learning Test-Revised (HVLT-R). Participants were healthy adults and patients diagnosed with probable dementia of the Alzheimer's type (DAT) or vascular dementia (VaD). Neuropsychological test records, which included the HVLT-R, were collected from an archival database. Ratio semantic clustering (RSC) scores were calculated for each HVLT-R recall trial. Factor analysis including RSC and other HVLT-R scores yielded high loadings of RSC scores on a distinct clustering factor. Group comparisons showed that normal controls had higher HVLT-R recall and RSC scores than DAT and VaD patients. Only the DAT group showed significant decline in RSC from the final learning trial to delayed recall, and only the DAT group performed more poorly than controls on other semantic processing measures (e.g., HVLT-R semantically-related false positives, Boston Naming Test). Results suggest that these newly developed semantic clustering indexes are easily calculated and potentially useful for discriminating between elder controls and dementia patients.