The Wisconsin Card Sorting Test (WCST) has been found to be valid in characterizing cognitive dysfunction in a variety of neurological disorders including traumatic brain injury (TBI). However, the test has been criticized as being too lengthy and frustrating for severely impaired patients. As a result, shortened versions of the WCST have been proposed. The utility of one of these shortened versions, the Wisconsin Card Sorting Test-64 (WCST-64), was examined in 106 persons with TBI. Findings showed strong associations between scores derived from the two tests. WCST scores were predicted from WCST-64 scores with sufficient accuracy for research purposes with adjusted R-squared values ranging from .74 to .87. Using the standard cutpoint of >40T to indicate impairment or normal performance for perseverative responses from each of the two tests, 91 (86%) of 106 subjects received the same classification showing substantial agreement (Kappa statistic 0.71; 95% CI 0.58-0.84). The WCST and the WCST-64 also performed comparably in predicting functional status at discharge from inpatient rehabilitation using the first score from a principal components analysis as a summary measure (both significant with p=.0002). These findings support the use of the WCST-64 in early evaluations of persons with moderate and severe TBI.