Abstract: Objective:The aim of this study was to test the repeatability of dolorimeter measurements in subjects with and without neck—shoulder symptoms and interobserver agreement in manual palpation. The second aim was to analyze how subjects, measurers, and the order of attempt influenced the repeatability of pain threshold (PT) measurements.Design:Repeated measurements in 100 female volunteers who were office workers.Setting:Interobserver repeatability was tested by examining 60 female office employees twice during the same day, and intraobserver repeatability was tested by examining 40 female office employees at an interval of 2 days.Patients:Their mean age was 38 (20–55) years; height was 163 (149–174) cm; and weight was 60 (44–115) kg.Main Outcome Measures:PT measurement by dolorimeter and manual palpation findings of four defined palpation points.Results:The inter- and intraobserver repeatability coefficients of the dolorimeter varied from 0.87 to 0.65; they were lower in subjects with neck-shoulder symptoms (DS) than in subjects with no or occasional symptoms (NOS). The sensitivity and specificity of dolorimeter for neck-shoulder symptoms was poor. The interobserver repeatability of manual palpation at trigger areas of trapezius and levator muscles varied from 0.15 to 0.62.Conclusions:The repeatability of dolorimeter is good, but sensitivity and specificity for neck-shoulder symptoms are poor. The dolorimeter is a good device for measurement of cervicobrachial tenderness when the subject acts as his or her own control. The repeatability of manual palpation is poor.