Objective: To investigate new aspects of the test-retest reliability of thermal quantitative sensory testing (QST) maxillofacial temperature in healthy young subjects, and differences between different parts and genders. Methods: Twenty-six healthy volunteers (14 women and 12 men) participated. Cold detection threshold (CDT), warm detection threshold (WDT), cold pain threshold (CPT) and heat pain threshold (HPT) were measured at three sites: the surface of left hand, and the left and right masseter. The testing was performed 3 times at 10 minutes intervals with each time of 3 stimulations at 4 seconds intervals, and repeated one week later. Data were analyzed with intraclass correlation coefficients (ICC) and four-way ANOVA for repeated measures of different genders, parts and time intervals. Results: Most variables showed acceptable to excellent reliability and non-significant difference across different test cycles (ICCCDT:0.642~0.869, ICCWDT:0.591~0.723, ICCCPT:0.672~0.967 and ICCHPT:0.757~0.917), and one-week intervals (ICCCDT: 0.508~0.772, ICCWDT: 0.560~0.885, ICCCPT: 0.629~0.872, and ICCHPT: 0.581~0.662). Bilateral contrast showed no statistical significance (PCDT=0.398, PWDT=0.223, PCPT=0.264 and PHPT=0.943). The CPT and HPT at the left hand were lower (lower sensitivity) than at the masseter (PCPT=0.003, PHPT=0.004). There were significant gender differences with higher sensitivity in women (PCPT=0.008 and PHPT=0.016). Conclusion: The reliability of most thermal threshold measures was acceptable for assessing somatosensory function. Facial skin is more sensitive to temperature-stimulated pain than the back of left hand skin, and female is more sensitive to temperature-stimulated pain than male.