The primary objective of this study was to provide contemporary normative data on aging and cognition from an ongoing community-based study. This dementia and stroke-free sample (age range = 20-79; mean = 53) consisted of 623 women and 322 men participating in the Maine-Syracuse Longitudinal Study at waves 4 to 6 (1993 to 2003). We employed a battery of 22 widely utilized cognitive tests. A 5 (age) x 3 (education) x 2 (gender) analysis of variance indicated that, in general, higher educated and younger participants exhibited better performance on cognitive tests. We found education group to be the strongest, and gender to be the weakest, predictor of cognitive performance. However, education cohort was not significantly associated with every cognitive outcome, nor was age cohort membership. The addition of cardiovascular disease health variables to a model including age, education, and gender groupings provided statistically significant, but modest, increases in prediction of performance on some tests. Results are discussed in relation to findings for previous studies presenting normative data on cognitive ability as a function of age, education, and gender.