The principle of rate-dependency has been proposed to explain the therapeutic effects of stimulant drugs in hyperactive children (HAC). This paper is a critical discussion of the salience of rate-dependency to childhood hyperactivity, on mathematical, theoretical and clinical levels. The results of a stimulant drug trial in 55 HAC are presented; the data are analyzed using analysis of variance to describe main drug effects, and these are compared to results derived from a traditional rate dependency analysis. The latter are found to have little salience to the actual clinical effects of stimulant drugs on a wide variety of behavioral, physiological and laboratory measures. The weakness of the rate dependency hypothesis, however, is not necessarily fatal to the idea that the state of the organism prior to drug administration influences the response profile of the drug. The heterogeneity of stimulant effects, and the relationship between stimulant effects and the predrug state of the organism, especially in electrophysiological paradigms, are clear. A hypothesis is presented to suggest that HAC may be characterized by a trait of excessive variability. Homeostatic stimulant effects in reducing response variability may be central to the therapeutic action of the drug. A neural substrate for the abnormal oscillations which characterize HAC, the correction of which is germane to therapeutic stimulant effects, is presented in terms of the regulatory functions of the frontal lobe. A neuroanatomic locus of childhood hyperactivity is proposed in terms of disorder or dysmaturation of frontal striatal systems.