A modified version of the Wisconsin Card Sorting Test (mWCST) proposed by Nelson (1976) was administered to 229 healthy community-dwelling older adults, composed of 97 men and 132 women, ages 45 to 91. Investigating the impact of demographic factors such as age, education, and gender on mWCST performance, results indicated that mWCST performance was significantly affected by both age and education. Unlike the standard WCST, however, gender did not significantly influence mWCST performance. Because demographic factors appear to exert a significant influence on mWCST performance of normal older adults, demographically-corrected norms were calculated according to the procedure described by Heaton, Grant, and Matthews (1991). In addition, longitudinal analysis of mWCST performance revealed that significantly fewer nonperseverative errors were committed at retest approximately one year later. Number of categories completed and perseverative errors did not appear to demonstrate significant practice effects in this sample. Accounting for demographic influences and the inspection of practice effects on serial administration of the mWCST may improve upon its sensitivity and specificity for use in the clinical assessment of executive function in older adults.