Abstract Objective The Boston Naming Test (Kaplan et al., 1983) is a 60-item test designed to measure confrontational naming ability. Word-finding difficulty is a prominent feature of aphasias, with picture-naming ability being a reliable predictor of lexical retrieval ability (Herbert et al., 2008). Lower performance on the BNT is found among older age groups and individuals with lower educational attainment (Zec et al., 2007). Notably, an item-level analysis of the BNT has not been analyzed in a Veteran population. There is great importance in properly accounting for demographic variables when interpreting patient test results. Methods An item-level confirmatory factor analysis examined mean BNT scores in a veteran sample (n = 517) ages 20–79 (M = 51) with years of educational attainment (M = 13). The purpose of this study was to conduct an item-level factor analysis of the unique cohort effects of both age and education on the Boston Naming Test within a Veteran population. Results The BNT solution explained 60.95% of the variance, and components reflected early and late trials. Within the rotated components matrix, education contributed to late items (36–60) whereas age contributed to early items (11–45). Conclusions These results indicate the importance of properly accounting for demographic variables such as age and educational attainment when interpreting test results unique to a Veteran population. Higher educational attainment is associated with increased performance on later items, whereas older age is associated with decreased performance on earlier items. It may be informative to consider item-level performance when assessing confrontational naming ability.