Objective Acute exercise can trigger a hypoalgesic response (exercise-induced hypoalgesia [EIH]) in healthy subjects. Despite promising application possibilities of EIH in the clinical context, its reliability has not been sufficiently examined. This study therefore investigated the between-session and within-subject test–retest reliability of EIH at local and remote body parts after aerobic cycling at a heart rate–controlled intensity. Methods Thirty healthy adults (15 women) performed 15 minutes of aerobic cycling in two sessions. Pressure pain thresholds (PPTs) were assessed at the leg (local), the back (semilocal), and the hand (remote) before, immediately after, and 15 minutes after exercise. Intraclass correlation coefficients (ICCs) were calculated for absolute and percent changes in PPT from baseline to immediately postexercise, and between-session agreement of EIH responders was examined. Results PPTs significantly increased at the leg during both sessions (all P < 0.001) and at the back during session 2 (P < 0.001), indicating EIH. Fair between-session reliability was shown for absolute changes at the leg (ICC = 0.54) and the back (ICC = 0.40), whereas the reliability of percent changes was poor (ICC < 0.33). Reliability at the hand was poor for both absolute and percent changes (ICC < 0.33). Agreement in EIH responders was not significant for EIH at the leg or the back (all P > 0.05). Conclusions Our results suggest fair test–retest reliability of EIH after aerobic cycling for local and semilocal body parts, but only in men, demonstrating the need for more standardized methodological approaches to improve EIH as a clinical parameter.