The relationship between impairment, measured by the Halstead-Reitan Neuropsychological Test Battery (HRNTB), and disability, measured by the Functional Independence Measure (FIM), was investigated in 164 subjects completing acute, inpatient rehabilitation following traumatic brain injury. Exploratory factor analysis of the FIM supported a two-factor model of disability, with 13 items loading on a motor disability factor and five items loading on a cognitive disability factor. HRNTB findings and injury-related variables were predictive of motor disability, as five variables accounted for 39-44% of the variance in FIM Motor scores. Contrary to expectations, the relationship with cognitive disability was comparable to, but did not exceed, that for motor disability, as only 29-40% of the variance in FIM Cognitive scores was accounted for. The WAIS-R Comprehension subtest was the single best predictor of cognitive disability, accounting for 20% of variance, and suggesting that the FIM Cognitive subscale is measuring social-cognitive ability, as intended by its authors. Results are discussed in terms of the relationships among impairment and disability, including the need to further scrutinize the operationalization of these constructs as they relate to cognitive functions.