We examined the effects of morphine microinjected into the medullary dorsal horn (MDH) on the ability of monkeys to detect temperature increases in the noxious heat range. Behavioral detection latency and the percentage of correct detections were used as measures of the perceived intensity of noxious heat stimuli. Three monkeys were trained to detect a change (T2) of 0.4, 0.6, or 1.0 degrees C from a previous noxious heat level of 46 degrees C (T1). Effects on attentional, motivational, and motoric aspects of the monkeys' behavior were assessed by having them detect innocuous cooling and visual stimuli in tasks of similar difficulty. Morphine (1, 3, and 10 micrograms) microinjected into the MDH produced a dose-dependent and stimulus-intensity-dependent increase in the latency to detection of the T2 stimuli. These effects were opiate receptor-mediated since they were antagonized by systemically administered naloxone (0.5 mg/kg, i.m.) given 40 min after the microinjection of morphine. There were no effects of morphine on the behavioral detection latencies to the innocuous cooling and visual stimuli, indicating that the effects of morphine were modality-specific and independent of changes in motivation, attention, or motoric ability. These data demonstrate a pharmacologically specific effect of opiates on the perceived intensity of noxious heat stimuli at the earliest central relay pathway transmitting noxious information.