Quantitative properties of the neural system mediating the rewarding and priming effects of medial forebrain bundle (MFB) stimulation in the rat have been determined by experiments that trade one parameter of the electrical stimulus against another. The first-order neurons in this substrate are for the most part long, thin, myelinated axons, coursing in the MFB and ventral tegmentum, with absolute refractory periods in the range .5-1.2 msec and conduction velocities of 2-8 m/sec. Local potentials in these axons decay with a time constant of about .1 msec. A supernormal period follows the recovery from refractoriness. These axons integrate current over exceptionally long intervals, accommodate slowly, and fire on the break of prolonged anodal pulses. These properties rule out the hypothesis that catecholamine pathways constitute the first-order axons. The second-order (postsynaptic) part of the substrate shows surprisingly simple spatial and temporal integrating characteristics. We examine the logic that permits conclusions of this sort to be derived from behavioral data and the role of these derivations in establishing neurobehavioral linkage hypotheses.In this article we describe properties of the neural tissue whose excitation eventuates in the reinforcing and motivating effects of electrical stimulation of the medial forebrain bundle (MFB) in the rat. The goal of the research is to identify these systems by anatomical and electrophysiological methods, thus providing physiological psychology with a model for studying the neurophysiological bases of learning and motivation in a higher vertebrate.The properties of the substrate for selfstimulation described in this article have been inferred from behavioral trade-off experiments-experiments that determine the value of one parameter of stimulation (e.g., current intensity) required to produce a criterion level of performance at each setting of another parameter (e.g., pulse duration). We review experiments of this kind in selfstimulation while examining two theoretical questions: (a) Why do behavioral trade-off functions have the power to reveal quantitative properties of the substrate for the behavior? and (b) Why must trade-off experiments play a pivotal role in relating anatomical and electrophysiological data to the behavioral phenomenon?Using microelectrodes, physiological psychologists have long been able to examine the response of single neurons to a behaviorally significant stimulus, such as rewarding brain stimulation (Rolls, 1975). The advent of 2-deoxyglucose autoradiography makes it possible to visualize a whole range of neural systems activated by such a stimulus (Figure 1). The existence of powerful techniques for directly revealing nervous activity confronts us with a conceptual problem that physiological psychology shares with other reductionist disciplines: How may one establish that a system defined by observations at one level of analysis explains phe-228