Background and purpose In animal studies, enhanced sensitivity to painful stimuli succeeding chronic stress has been reported, while acute stress is reported to induce analgesia. Human studies on the effect of mental stress on pain are more equivocal. A disturbed stress-response resulting in an increased sensitivity to painful stimuli has also been discussed as a potential mechanism for e.g., the fibromyalgia syndrome. Endogenous analgesia may be studied in humans by measuring the analgesic effect of heterotopic noxious conditioning stimulation. In neurophysiological animal studies this phenomenon was originally denoted "diffuse noxious inhibitory controls" (DNIC), but for human studies it has been suggested to use the term conditioned pain modulation (CPM). The clinical relevance of aberrances in CPM is not clear. Inhibitory CPM is reported as being reduced in several medically unexplained syndromes with musculoskeletal pain aggravated by mental stress. However, whether the reported reduced CPM effects are causally related to clinical pain is unknown. In the present study the effect of a mental stressor on CPM is studied. Methods With tourniquet-induced pain as the conditioning stimulus we estimated the CPM effect in twenty healthy subjects. Heat pain threshold (HPT), supra-threshold heat pain level (SHPL) and pressure pain threshold (PPT) were used as test stimuli. Measurements were performed at baseline, after a stressful task and after a non-stressful task presented in a blinded cross-over design. We used repeated-measures ANOVAs in the analysis with simple contrasts for post hoc analysis. Results With a ANOVA repeated measures model we found a significant task effect (F = 18.5, p ≤ 0.001), indicating that CPM was successfully induced. In our ANOVA model, we found a significant effect of stress in the contrast analysis (F = 5.2, p = 0.037), indicating that CPM was affected by the stressful task. The effects on PPT could not be analyzed due to a significant carry-over effect (for PPT only). Conclusions In the present blinded crossover study, we found a significant small to medium inhibitory effect of mental stress upon the CPM of thermal pain. Implications Our results suggest that previously reported reduced inhibitory CPM in several medically unexplained syndromes with musculoskeletal pain aggravated by mental stress possibly can be related to confounding or clinically relevant stress level differences. However, the result might be modality-specific. Further studies in patients are obviously needed, and the impact of mental stress on CPM should be investigated also with other stressors.