According to the 2011 ICRP statement, the threshold in absorbed dose for the lens of the eye is now considered to be 0.5 Gy. Therefore, it is important to keep the eye lens doses during head computed tomography examination as low as reasonably achievable. In this study, the two lenses and the occiput doses associated with patient setup errors and head size were evaluated using phantoms developed for pediatric patients. The phantoms were made using flexible acrylic sheets in a cylindrical shape [diameters of 6 (premature baby), 8 (neonate), 10 (infant), and 12 cm (child)] and placed on the end of the CT bed. The bed position was raised from the center for each phantom size. The two lenses and the occiput doses at each setup were measured using radiochromic film. By raising the bed position, the two lenses doses decreased at all phantom sizes. However, the occiput doses changed in a complex manner according to the phantom size. In this study, decreasing the distance from the lens to the X-ray tube (i.e. raising the bed position) is an effective way to reduce the absorbed dose of the lens. However, when the occiput was positioned at the center of the gantry aperture, the occiput dose peaked in all phantom sizes. In addition, the results indicated a complex dose distribution in pediatric head CT when taking into account the bed position (patient setup errors) and phantom sizes. Furthermore, beam slice width and pitch are important factor in determining the maximum absorbed dose. To keep radiation doses during CT examination as low as reasonably achievable, it is important to clarify the influence of CT scanning settings on CT dose.