The aim of this study was to evaluate the intra- and inter-examiner reliability and validity of neck range of motion (ROM) measurements. Thirty-five healthy subjects were assessed in all neck movements from two initial positions, sitting and standing, actively (open and closed eyes) and passively by using a 3D ultrasound-based motion analysis device (Zebris). Three tests were employed to assess intra-examiner reliability and two examiners used for the inter-examiner reliability. X-rays in neck flexion and extension were used to validate the Zebris system. The standing position yielded higher intraclass correlation coefficient (ICC) values (>0.86) with less error [smallest detectable difference (SDD)<13.8%] than sitting (ICC>0.79, SDD<14%). Passive assessment of neck ROM presented better reproducibility than active assessment with open or closed eyes in both positions. The inter-examiner reliability was moderate (ICC=0.43-0.68). The correlation between the Zebris system and X-rays was high in both flexion and extension movements. The results showed that the most reliable protocol for assessment of neck ROM is a passive measurement in the standing position. The measurements were well validiated against X-rays and the experience of the investigators must be considered before any comparison among studies is employed.