Most investigators now agree on the existence of abnormalities of brain function in children with learning disabilities with or without hyperactivity. However, attempts to define the neurobiological substrates have been hindered by many factors. On the clinical level, learning disability is not a monolithic entity but an operational term used to define a number of specific conditions. The only cornmod feature is difficulty with school performance in the face of normal intelligence. Until homogeneous subgroups can be defined, there is no good reason to suspect a single, discrete biological marker.Neuropathological studies are limited to 2 cases. Attempts to visualize the brain by computerized tomography have provided limited but inconsistent evidence of variations from normal hemispheric widths and asymmetries. Neurophysiological studies have succeeded in separating learning disabled children from normal controls by the application of discriminant analysis and other statistical techniques. Computer-assisted analysis of brain bioelectrical activity has sharpened these tools. Unfortunately, the techniques have not been able to narrow the focus to single, discrete cerebral areas; rather, they have shown dysfunction in widely separated but probably interconnected regions.The only neurochemical studies available have been performed on hyperactive children rather than on patients with uncomplicated learning disabilities. A hypothesis that the underlying pathophysiology involves dopamine depletion has the most support and provides an explanation for some of the therapeutic effects of psychostimulant drugs. Problems are the focus on motor aspects of hyperactivity rather than the attention deficit and heavy dependence on the use of clinical rating scales. The biochemical analyses and conclusions would be more powerful if better psychophysiological markers could he developed, including such factors as attentiveness, distractability, vigilance, attention span, and reaction time.The results of physical and neurological examinations as well as of clinical genetic studies also point toward a biological basis for learning disabilities. These clinical tools help to sharpen our focus but still provide insufficient specificity. Productive search will require close collaboration between neuropsychologists and investigators using the quantitative tools of neuroanatomy, neurophysiology, and neurochemistry.Golden GS: Neurobiological correlates of learning disabilities. Ann Neurol 12: [409][410][411][412][413][414][415][416][417][418] 1082 The child who does not meet expectations for academic performance in school but has intelligence in the normal range has been a subject of research for many years. Concepts concerning these children have been in constant evolution. Initially, especially if the child was from a middle or upper class socioeconomic background, attempts were made to ascribe the difficulty to lack of motivation or to a "block" in the learning process due to psychological dysfunction [30]. The child was then offered psyc...