1. Psychophysical experiments were initiated to determine the possible influence of increasing stimulus size on perceived pain intensity. Six trained human subjects (5 male, 1 female) made visual analogue scale (VAS) ratings for pain-sensation intensity and unpleasantness in response to nociceptive thermal stimuli. Test stimuli consisted of 5-s duration heat pulses (45-50 degrees C in 1 degrees increments) delivered by one, two, or three contact thermal probes (1 cm2 each) applied to the medial aspect of the anterior forearm. 2. The area of skin receiving noxious thermal stimuli was changed by randomly varying the number of thermodes activated. The effects of varying the distance between the thermal probes also were evaluated. In the first series of experiments, thermal-probe separation was kept close to 0; in subsequent experimental series, the thermodes were separated by either 5 or 10 cm. 3. In each experimental series, considerable spatial summation occurred in both pain-sensation intensity and unpleasantness dimensions of pain. This summation occurred throughout the nociceptive thermal range of 45-50 degrees C and was larger at suprathreshold temperatures (greater than or equal to 47 degrees C) than those near threshold (less than or equal to 46 degrees C). Unlike spatial summation of perceived warmth, that of pain was not characterized by systematic changes in power-function exponents but as approximately upward parallel displacements in double-logarithmic coordinates. 4. Thermal-probe separation over a range of 0-10 cm had no effects on spatial summation of pain-sensation intensity or pain unpleasantness. In contrast, increasing thermal-probe separation increased the subjects' ability to discriminate differences in stimulus size and their ability to detect correctly the number of thermal probes activated. 5. Because affective VAS ratings of unpleasantness were linearly related to, but distinctly and systematically less than, VAS ratings of pain-sensation intensity, it was clear that subjects responded quite differently to these two pain dimensions. Affective judgements were not additionally influenced by thermal probe separation and hence by the ability to perceive stimulus size or number of thermal probes activated. 6. The results indicate that powerful spatial-summation mechanisms exist for heat-induced pain. Spatial summation of pain is likely to be subserved both by local integration mechanisms at the level of single spinothalamic-tract neurons and by recruitment of central nociceptive neurons, because spatial summation of pain occurred to approximately equal extents under conditions of thermode separations over a distance of at least 20 cm.