Cognitive neuropsychology's domain of inquiry concerns the structure of normal perceptual, motor, and cognitive processes. As such, it constitutes a branch of cognitive science. Cognitive neuropsychology differs from other branches of cognitive science only by the type of observation that it uses in developing and evaluating theories of normal cognition. The data used in cognitive neuropsychology are the patterns of performance produced by brain-damged subjects. Because the basic data used in cognitive neuropsychology are the result of a biological manipulation—a brain lesion—these data will be relevant to claims about the functional organization of the brain. Hence, cognitive neuropsychology may also be considered to be a branch of cognitive neuroscience. However, in this paper I will be concerned with an assessment of research programs whose principal or only aim is to constrain theories of normal cognitive functioning through the analysis of acquired disorders of cognition. Following a brief discussion of the basic assumptions that motivate cognitive neuropsychological research, I consider Kosslyn and Van Kleeks (1990) claim that the study of brain-damaged subjects for the purpose of constraining theories of normal cognitive processing cannot lead to meaningful conclusions unless the theories are directly cast in terms of anatomical and physiological facts. I argue that these authors conflated criticisms that may apply to any empirical science with criticisms that may apply specifically to cognitive neuropsychology. Separate consideration of the criticisms specific to cognitive neuropsychology reveals that these are unfounded. The main point of this discussion is to emphasize the pragmatic character of the motivation for using impaired performance to constrain theories of normal cognition. The usefulness of cognitive neuropsychological research is illustrated through specific examples.