Narcotics reduce the behavioral response to injury, while leaving unaffected the response to other stimuli. This effect could be due to a reduction of the central transmission of injury signals (modulation of sensory throughput) or to an effect on higher inte~ ative levels concerned with stimulus recognition or stimulus interpretation**. Since much of the functional capacity of the central nervous system and nearly every level of the neuraxis is involved in signal processing, observations limited to the effect of systemically administered narcotics on behavior will provide little information regarding the specific sites and mechanisms of action of narcotic analgetics. Investigators interested in these aspects of narcotic activity have therefore resorted to experimental approaches which permit a detailed study of the interaction between narcotic analgetics and particular regions and/or functional systems within the central nervous system. Recording of local electrical activity following systemic application of a narcotic does not specify where the narcotic is acting. Similarly local lesions or electrical stimulation can only indicate through which systems narcotics may act and do not reveal the causal chain between injury and pain. As a consequence of the inadequacies of these approaches, investigators have turned with increasing frequency to methods which permit the application of drugs directly to single neurons or circumscribed neuronal pools. The mi~ roiontophoretic technic has provided much of our limited information concerning the direct effects of narcotic analgetics on neuronal activity.~ n~ o~ unately, the microiontophoretic stimulation of a single neuron usually is incapable of eliciting a detectable behavioral response. Therefore, even though many neurons can be tested for their individual sensitivities to suspected neurotransmitter or other agents, it cannot be determined by ionto-