Correct diagnosis and analysis of an abnormal head posture (AHP) is critical in the clinical assessment of patients with strabismus and nystagmus. This study was performed to evaluate the accuracy and precision of clinicians estimating the degree of face turns and head tilts.Experimental simulation study.Two normal volunteers and 4 pediatric and 1 general ophthalmologists.Using a cervical range of motion (CROM) device, single axis head position (chin up, chin down, right and left face turn, and head tilts) were modeled, and clinical estimates of the amount of deviation were estimated by the four ophthalmologists. Each of the 4 examiners estimated 108 single axis AHPs in a randomized and masked pattern.Accuracy of clinical estimation of AHP, intra- and inter- observer variability, and variability in terms of head position.The examiners tended to overestimate the AHPs by an average of 5 degrees (P less than 0.001). Two "right" positions (right head tilt and face turn) were the worst. The average error was 10 +/-8 degrees for the right tilt and turn positions, while the error for all other positions, excluding these two, averaged 0 +/-10 degrees (P = 0.0001).Clinical estimation of an AHP by ophthalmologists is in part remarkably accurate, even if imprecise, but some positions, notably right head tilt and right face turn in this study, were significantly overestimated. Objective methods of measuring AHPs such as the CROM may be preferable for these RHT and RFT AHPs and when maximum precision is needed for tracking changing AHPs or deviations.