In laboratory experiments with hyperkinetic, untrainable dogs and in a comparison of levoamphetamine, dextroamphetamine, and placebo in children, levoamphetamine and dextroamphetamine were found to be approximately equal in calming an aggressive, hostile dog and in benefiting "unsocialized-aggressive" children; dextroamphetamine was more effective than levoamphetamine in calming "nervousness" and hyperactivity in dogs and in overanxious-hyperkinetic children. These data suggest that in the hyperkinetic syndrome, aggression and hostility may be benefited equally by levoamphetamine or dextroamphetamine via a dopaminergic mechanism, while anxiety and overactivity may be benefited significantly only by the dextro isomer via a norepinephrinergic mechanism.The plight of children with behavior and learning problems that often respond to stimulants by "paradoxical" calming has become well known under such names as "hyperkinetic syndrome," "minimal brain dysfunction," and "minimal brain damage." These and other grammatically singular appellations are bandied as if we believed these problems represented a single, homogeneous clinical entity (like pneumococcal pneumonia) that would uniformly respond to a prescribed treatment, if only we could find the treatment. However, those familiar with the syndrome readily admit its diversity of clinical pictures, probable diversity of etiology and pathogenesis, and even diversity of response to the highly touted stimulant therapy.At least two authors have recently attempted to clean up this diagnostic cesspool. Fish (1) has proposed criteria by which children who may respond favorably to stimulants could be assigned to one of three diagnostic categories: 308.0-hyperkinetic reaction (with immaturity, inadequacy, lability, and poor organization); 308.2-overanxious reaction ("nervousness" and possible overactivity with subjective distress, otherwise well-organized behavior); and 308.4-unsocialized-aggressive reaction (aggressiveness and hostility with denial of feelings and personal responsibility, otherwise well-patterned behavior). These categories are reminiscent of Wender's "classical hyperactive," "neurotic," and "sociopathic" subsyndromes (2).One of the puzzling things about the hyperkinetic syndrome that raises serious questions about its being a homogeneous diagnostic entity is the individual variation in response to stimulants. It has long been known (3, 4) that sometimes dextroamphetamine helps a hyperkinetic child more than the racemic (dl) mixture does, and sometimes vice versa. In addition, some children are helped by methylphenidate (Ritalin) but not by amphetamine, and vice versa. Furthermore, while the majority of such children are helped by stimulants, a small percentage are actually made worse. The uncertainty concerning the practical clinical use of drugs for these children is underscored by recent attempts to introduce some prognostic guidelines. Barcai (5) has proposed a finger twitch test. Yoss and Moyer (6) have found that 25 percent of hyperkinetic and underachie...