1. We wished to determine whether there are differences in the spatial pattern and intensity of synaptic activity within the conscious human forebrain when different forms and intensities of innocuous and noxious thermal stimuli are experienced. Accordingly, positron emission tomography (PET) with intravenous injection of H2(15)O was used to detect increases in regional cerebral blood flow (rCBF) in normal humans as they discriminated differences in the intensity of noxious and innocuous thermal stimulation applied to the nondominant (left) arm. After stereotactic registration, subtraction images were formed from each subject by subtracting counts of emissions obtained during lower-intensity stimulation from those obtained during stimulation at higher intensities. A statistical summation analysis (Z score) of individual voxels was performed. In addition, volumes of interest were chosen on the basis of a priori hypotheses and the results of previously published PET studies. In both types of analysis, statistical thresholds were established with corrections for multiple comparisons. 2. Twenty-seven subjects were divided into three groups of nine subjects each for the three phases of this investigation. For studies in which repetitive contact heat stimuli were used, each subject was instructed in magnitude estimation on the basis of a scale for which 0 indicated "no heat sensation," 7 "just barely painful," and 10 "just barely tolerable." For the study of pain elicited by immersion of the hand in cold water, subjects were instructed to use a scale in which 0 represented "no pain" and 10 represented just barely tolerable pain. 3. In the warm-discrimination study, two intensities of innocuous heat (36 and 43 degrees C) were applied with a thermode as repetitive 5-s contacts to the volar forearm for a total of approximately 100 s, 8 stimuli before and 12 during each scan. Each temperature was applied on alternate scans for a total of four scans per subject. Neither stimulus was rated painful. All subjects discriminated the 43 degrees C stimulus (average rating 5.90 +/- 1.43, mean +/- SD) from the 36 degrees C stimulus (1.96 +/- 1.08, mean +/- SD; t = 13.19, P < 0.0001). Significant increases in rCBF to the 43 degrees C stimuli were found in the contralateral ventral posterior thalamus, lenticular nucleus, medial prefrontal cortex (Brodmann's areas 10 and 32), and cerebellar vermis. 4. The procedure for discriminating between noxious and innocuous heat stimuli was identical to that used for warm discrimination except that the stimulation temperatures were 40 and 50 degrees C. All subjects rated the 50 degrees C stimuli as painful (average rating 8.9 +/- 0.9, mean +/- SD) and the 40 degrees C stimuli as warm, but not painful (2.1 +/- 1.0). Significant rCBF increases to 50 degrees C stimuli were found contralaterally in the thalamus, anterior cingulate cortex, premotor cortex, and secondary somatosensory (S2) and posterior insular cortices. Significant activity also appeared within the region of the contralateral anterior insula and lenticular nucleus. The ipsilateral premotor cortex and thalamus, and the medial dorsal midbrain and cerebellar vermis, also showed significant rCBF increases. Cerebral blood flow (CBF) increases just below the threshold for statistical significance were seen in the contralateral sensorimotor cortex [primary motor cortex (M1)/primary somatosensory cortex (S1)]. 5. For discrimination between tonic innocuous cold and tonic cold pain, the left hand was immersed to the wrist, throughout each of six scans, in water kept at an average temperature of either 20.5 +/- 1.15 degrees C (mean +/- SD) or 6.02 +/- 1.18 degrees C (mean +/- SD) on alternate scans. All subjects rated the intensity of the stimuli on a scale in which 0 indicated no pain and 10 represented barely tolerable pain. Subjects rated the 20 degrees C water immersion as painless (average rating 0.18 +/- 0.48, mean +/- SD), but gave ratings indicating i