Purpose Heterophoria is the misalignment of the eyes in monocular viewing and represents the accuracy of vergence driven by all classical cues except disparity. It is challenging to assess restless children using clinical cover tests, and phoria in early childhood is poorly understood. Here we used eye tracking to assess phoria as a function of viewing distance and target in adults and young children, with comparison to clinical cover tests. Methods Purkinje image tracking (MCS PowerRefractor) was used to record eye alignment in adults (19–28 years, N = 24) and typically developing children (three to five years, N = 24). Objective unilateral and alternating cover tests were performed using an infrared filter while participants viewed a pseudo-randomised sequence of Lea symbols (0.18 logMAR; Snellen: 20/30 or 6/9) and animated cartoon movies at distances of 40 cm, 1 m, and 6 m. For the unilateral cover test, a 10 second binocular period preceded and followed 30 seconds of occlusion of the right eye. For the alternating cover test, a 10 second binocular period preceded and followed alternate covering of right and left eyes for three seconds each. Phoria was derived from the difference in weighted average binocular and monocular alignment. A masked prism-neutralised clinical cover test was performed for each of the conditions for comparison. Results Closer viewing distance resulted in greater exophoria for both children and adults (p < 0.001). Phorias were similar for adults and children for each viewing distance and target, with mean differences of less than 2 prism dioptres or pd. Overall, the average PowerRefractor phorias in pd (pooled across protocols) for adults were 1.3, 2.3 and 3.8 pd exophoria and for children were 0.1 pd esophoria, 0.94 and 3.8 pd exophoria for the 6 m, 1 m and 40 cm distances respectively. The corresponding clinical cover test values were 0.7, 1.9, and 4.1 pd exophoria for adults and 0, 1.5 and 3.3 pd exophoria for the children. Refractive states were also similar (≤ 0.5 D) for viewing the Lea symbols or movie for any protocol tested. Conclusions Phoria estimation can be challenging for a pre-school child. These data suggest that by three to five years of age objective eye-tracking measures in a typically developing group are adult-like at the range of distances tested, and that use of an animated movie produces similar average results to a small optotype (0.18 logMAR; Snellen 20/30 or 6/9).