The relationships among neuropsychological assessment, situational assessment, and judgments of future employability in 127 survivors of moderate to severe closed head injury were evaluated. Participants received a comprehensive neuropsychological battery and a situational vocational evaluation. Two groups were created, based on employment recommendation from the situational evaluation. A seven-factor model was found to account for the preponderance of variance within the neuropsychological battery used, and one factor was predictive of group assignment from situational assessment. However, the present data reinforce the predominance of findings in the literature that indicate that, in general, no individual neuropsychological test can be used to predict vocational performance in more environmentally relevant environments. Exceptions to this "rule" may occur when comparisons between results of formal neuropsychological tests and situational evaluation are limited to very simple, very circumscribed, and/or very well-defined functions. Thus, situational assessment is seen as a critical adjunct to neuropsychological assessment in making "real-world" predictions. The particular situational assessment used in this study was internally valid, an important finding given the importance of situational assessment in ecologically valid predictions.