Little information is available concerning the influence of subconscious mechanisms on neuroendocrine function, more specifically, proenkephalin peptide F release. Ten men [5 middle distance runners (21.6 (SD 0.54 years) and 5 untrained men (24.0 (SD 4.3 years)] consented to be volunteers in this investigation. Submaximal exercise intensities of 25% and 50% of peak oxygen consumption (VO2) (8 min stages) were used for both the control and hypnosis treatments. A traditional hypnotic induction was used, with the suggestion of two higher intensities of exercise stress (50% and 75% peak VO2) previously experienced in familiarization and testing by each subject. Each minute oxygen consumption was measured using open circuit spirometry, heart rate via an ECG, and ratings of perceived exertion (RPE) using the Borg scale. Plasma peptide F immunoreactivity (ir) [preproenkephalin-(107-140)] in blood sampled from an indwelling cannula was measured by radioimmunoassay at 7-8 min of each stage of the exercise test. Expected significant increases were observed for all cardiorespiratory and perceptual variables over the increasing exercise intensities and there were no significant differences between trained and untrained groups for peptide F if response patterns. Hypnosis did not significantly affect peptide F ir concentrations (P > 0.05) and did not significantly alter exercise heart rate, RPE or minute ventilation (P > 0.05). However, hypnosis did significantly increase oxygen consumption during exercise (P = 0.0095) but not of the magnitude needed for the metabolic demands of the higher exercise intensities. Thus, traditional hypnosis was unable to make functionally significant changes in the cardiorespiratory variables.(ABSTRACT TRUNCATED AT 250 WORDS)