The present study examined methods of control- and experimental-subject selection in neuropsychological research. Using both meta-analytic and prospective procedures, it was found that varied subject-selection procedures can potentially confound the results of neuropsychological studies and, by extension, can thwart the formulation of laws regarding neuropsychological function. In terms of selecting control subjects, the current findings suggest that individuals who self-report a negative history for neuropathology and psychopathology are adequate, perhaps even preferable, for inclusion. In terms of selecting experimental subjects, it appears preferable, if not essential, to obtain lesion confirmation through relevant neurodiagnostic tests. Based on these findings, it is recommended that subject selection be made on the basis of scientific merit rather than convenience or availability. Published by Elsevier Science Ltd