We compared the sensitivity to brain dysfunction of an ability focused neuropsychological battery (AFB), as a proxy for the core of a flexible battery, to the Halstead-Reitan Battery (HRB). The AFB was designed to represent constructs of language function, fine motor skill, working memory, processing speed, verbal and visual memory, and verbal and visual abstraction and problem solving. Receiver operating characteristic analysis (ROC) yielded an area under curve (AUC) of. 86 for the AFB, versus. 83 for the HRB (p =. 50), for discriminating 54 patients with brain dysfunction due to various etiologies, from 69 non-neurologic medical controls. Additionally, Bayesian Model Averaging selected four tests from the combined set of AFB and HRB subtests, plus Trail Making B, which optimally discriminated the brain dysfunction from medical control patients: H-Words, Grooved Pegboard, Finger Tapping, and Trail Making B. These data support the current mainstream practice in neuropsychology of using an AFB (flexible battery) to assess brain dysfunction. In particular, tests involving processing speed appear to be among the most sensitive measures of brain dysfunction. The data do not support the superiority of the HRB to AFB approaches.