The purpose of the study was to evaluate the accuracy and precision of a rigid positioning device for repositioning the cervical spine accurately and precisely during conformal radiation therapy of dogs. Fifteen purpose bred research dogs in a radiation therapy study were included. The dogs were positioned using a head holder and a deflatable pillow attached to the treatment table. Port films were reviewed retrospectively, and repositioning precision was recorded by measurements in three orthogonal planes of the head, 2nd cervical vertebra and 1st thoracic spinous process. Mean treatment position was compared to the planning position for a measurement of systematic set-up error. Mean interfraction position variation of the 2nd cervical vertebra was 0.2, 0.1 and 0.2 cm for the ventrodorsal, caudocranial and laterolateral directions respectively, and the average systematic set up error was 0.2, 0.1 and 0.2 cm for the ventrodorsal, caudocranial and laterolateral directions respectively. Knowledge of the magnitude of reposition errors should be included when determining the margins around the tumor.