Objective To study an objective refraction method that includes different models of higher order aberrations; to study the relationship between manifest and objective refractions from wavefront aberrations with different pupils sizes.Methods Sixty-two eyes were included in this crosssectional study.Wavefront aberrations were measured using a Shack- Hartmann aberrometcr (WaveScan wavefront system, VISX, Inc., Santa Clara, CA). Refractive errors were computed that included different models of higher order aberrations for different pupil sizes. These refractions were then compared to manifest refractions using the Friedman test.Results The mean differences in refractive errors from wavefront aberrations were smallest between manifest refraction and objective refraction when only lower order aberrations were included.For 2,3,4,5,and 6 mm pupil sizes,the refractions for the median sphere were -0.540,-0.473,-0.422,-0.407 and -0.290 D,respectively.The differences were statistically significant (Z=99.29,P＜0.01).The refractions for the median cylinder were 0.193,0.142,0.119,0.078 and 0.065 D,respectively.The differences were statistically significant (Z=30.25,P＜0.01).The parameters for the median axis were -2.41,-2.89,-3.03,-1.94 and -2.40°,respectively.The differences were not statistically significant for all pupil sizes (Z=4.42,P＞0.05).Refractive errors with higher order aberrations included were significantly different from manifest refractions.Conclusion Objective refraction from wavefront aberrations varies with pupil size when models for higher order aberrations are included.The equivalent sphere is smaller when the spherical aberration is included.The refractive errors for lower order aberration for larger pupil diameters is in better agreement with manifest refraction than smaller pupil sizes.The mean axis difference is small between the two methods.