The use of quantitative sensory testing (QST) has become more widespread, with increasing focus on describing somatosensory profiles and pain mechanisms. However, the reliability of thermal QST has yet to be established. We systematically searched the literature using key medical databases. Independent reviewers evaluated reliability data using the Quality Appraisal for Reliability Studies checklist. Of the 21 studies we included in this review, we deemed 5 to have high methodological quality. Narrative analysis revealed that estimates of reliability varied considerably, but overall, the reliability of cold and warm detection thresholds ranged from poor to excellent, while heat and cold pain thresholds ranged from fair to excellent. The methodological quality of research investigating the reliability of thermal QST warrants improvement, particularly in terms of appropriate blinding. The results from this review showed considerable variability in the reliability of each thermal QST parameter.