Background and aims Exercise is commonly used as treatment for chronic pain with positive long-term effects on pain and pain-related disability. In pain-free subjects, hypoalgesia following an acute bout of exercise compared with a control condition has consistently been demonstrated also known as exercise-induced hypoalgesia (EIH). Walking exercise, a low intensity aerobic exercise, is frequently used in clinical practice as an easily applicable intervention for patients with chronic pain. Walking exercise is furthermore recommended as an effective treatment for patients with chronic musculoskeletal pain conditions to alleviate pain and reduce disability, however, the effect of walking on pain sensitivity is currently unknown. The aims of the present study were to investigate (1) the acute effect of walking on pain sensitivity, and (2) the relative (between-subjects) and absolute (within-subject) test-retest reliability of the hypoalgesic response across two sessions separated by 1 week. Methods In this randomised experimental cross-over study including two identical sessions, 35 pain-free subjects performed a standardized 6 min walking test and a duration-matched quiet rest condition in a randomized and counterbalanced order in each session. Before and after both conditions, handheld pressure pain thresholds (PPTs) were assessed at the thigh and shoulder, and pressure pain thresholds (cPPT) and pain tolerance (cPTT) were assessed with computer-controlled cuff algometry at the lower leg. Change in the pain sensitivity measures were analysed with repeated-measures ANOVAs, and test-retest reliability with intraclass correlation coefficients (ICC) and agreements in classification of EIH responders/non-responders between the two sessions. Results All subjects completed the walking conditions in both session 1 and session 2. The perceived intensity of walking assessed with rating of perceived exertion (RPE) and walking distance did not differ significantly between session 1 (distance: 632.5 ± 75.2 meters, RPE: 10.9 ± 1.9) and session 2 (distance: 642.1 ± 80.2 meters, RPE: 11.0 ± 2.4) (p > 0.11). Moreover, RPE showed excellent relative reliability with an ICC value of 0.95 [95%CI: 0.90–0.97]. Walking increased pain tolerance (mean difference: 2.6 kPa [95%CI: 0.5–4.9 kPa; p = 0.02]), but not pain thresholds compared with rest in both sessions. Hypoalgesia after walking demonstrated fair to good relative reliability (ICC = 0.61), however the agreement in classification of EIH responders/non-responders (absolute reliability) across sessions was low and not significant (κ = 0.19, p = 0.30). Conclusions Walking consistently increased pain tolerance but not pain thresholds compared with a duration-matched control condition with fair to good relative reliability between sessions. Based on classification of EIH responders/non-responders the absolute reliability between the two sessions was low indicating individual variance in the EIH response. Future studies should investigate the hypoalgesic effect of a walking exercise in a clinical pain population.