Objective:To observe the reliability and validity of using a head-neck relocation test (HRT) in assessing cervical position sense.Methods:A total of 62 subjects were recruited, including 32 with neck pain and 30 healthy subjects. All received two HRTs conducted by two independent assessors at a first session, and then they received another HRT 24 hours later administered by one of them. The joint position error (JPE) was measured in the four directions of cervical motion: flexion, extension, levorotation and dextrorotation. The intra- and inter-rater reliability was quantified in terms of intra-class correlation coefficients (ICCs) and the difference in JPE between the two groups was also determined for validity.Results:The intra- and inter-rater reliability for the HRT in the four directions had ICC values of 0.603-0.830 and 0.597-0.837 respectively. The corresponding standard error of measurement (SEM) values were 1.52-2.15cm and 1.69-2.09cm, while the minimal detectable change (MDC) values were 4.20-5.97cm and 4.69-5.79cm, respectively. The patients with neck pain had significantly larger JPE in cervical extension and right rotation compared to healthy individuals, on average. And their JPE for cervical extension had an area under the curve (AUC) of the receiver′s operating curve of 0.755.Conclusions:The HRT showed good intra- and inter-rater reliability in assessing cervical joint position sense. The JPE for cervical extension can be used to discriminate those with neck pain from healthy subjects.