The purpose of this study was to estimate inter-examiner reliability of head and neck algometry. Pain perception thresholds were assessed with a mechanical pressure algometer in 21 healthy individuals. Thresholds were assessed at 13 symmetrical points on each side of the head and neck, at the deltoid muscle and at the median finger. The pressure range of the instrument proved insufficient to study the pain perception threshold on the finger, however. Two different examiners carried out one or two examinations in each subject during one day. The sequence of investigations was varied randomly. The inter-examiner reliability was found to be good, with a mean intra-class correlation coefficient (ICC) of 0.75. Intra-examiner reproducibility was excellent (mean ICC = 0.84). The mean inter-examiner coefficient of variation was 18.7%, while the mean coefficient of repeatability (CR) was 1.60 kg/cm2. In comparison, the mean intra-examiner coefficient of variation was 15% while the mean CR was 1.29 kg/cm2. Statistically significant differences between examiners were found for the frontal point (p < 0.01), while a trend towards lower thresholds in one of the two observers was seen in 10 of the 13 non-significant points. Inter-examiner reliability of side differences was excellent, with CR = 1.23 kg/cm2. In conclusion, manual algometry with a rather inexpensive mechanical device has a good to excellent inter-rater reliability. When studying patients, however, the possible bias introduced by different examiners should be taken into account, both regarding study design and data analysis.