1. The dose-related effects of azapropazone on (i) event-related and spontaneous EEG-activity and (ii) the subjects' pain ratings were investigated using an experimental human pain model based on both chemo-somatosensory event-related potentials (CSSERP) and subjects' pain ratings. 2. Healthy subjects (n = 20) participated in a placebo-controlled, randomized, double-blind, four-way cross-over study. Single doses of azapropazone (300 mg, 600 mg and 1200 mg) and placebo were administered intravenously. Each experiment consisted of five sessions (before and 1, 2, 4 and 8 h after administration of the medication). Each session lasted for approximately 40 min. In the first 20 min, pain was induced by short CO2-stimuli presented to the right nostril (phasic pain; interstimulus interval 30 s) and EEG was recorded from five positions. CSSERPs were obtained in response to painful CO2-stimuli. In the following 20 min period, tonic pain was induced by a constant stream of dry air introduced in the left nostril. Subjects rated the intensity of both phasic and tonic pain by means of a visual analogue scale. Additionally, a frequency analysis of the spontaneous EEG was performed. 3. Azapropazone reduced the pain-related CSSERP-amplitudes at frontal and parietal recording positions. This topographical pattern was observed in previous studies with opioids, while NSAIDs such as flurbiprofen and ketoprofen exerted effects at frontal and central positions. In contrast to other NSAIDs, administration of azapropazone resulted in a reduction of the frequency bands alpha 1, delta and theta of the spontaneous EEG. At the subjective level, analgesic effects of azapropazone were observed in the ratings of tonic pain. 4. Analgesic properties of azapropazone were demonstrated in man. The topographical pattern of the changes in the CSSERPs and the effects on EEG background activity suggest a central component of the analgesic action of azapropazone.