Purpose: To determine the correlation between the refractive and measured corneal power changes after myopic photorefractive surgery.Setting: Department of Ophthalmology, National Taiwan University Hospital, Taipei, Taiwan.Methods: Eighty-six eyes that had myopic photorefractive surgery were analyzed. The data included preoperative and 1-year postoperative subjective refraction, standard automated keratometry, and computerized videokeratography. Statistical analysis was performed to determine the relationship between the changes in subjective refraction in the corneal plane (ΔSEQco) and in 4 corneal power measurements including the power measured by automated keratometry (ΔAuto K), topographic-simulated keratometric power (ΔSim K), the power of the first photokeratoscopic ring on videokeratography (ΔCentral K), and the average videokeratographic power on the pupil margin (ΔPupil K).Results: The measured corneal power always underestimated the ΔSEQco, with ΔSEQco > ΔCentral K > ΔSim K > ΔPupil K > ΔAuto K. All the changes in measured corneal power could predict the ΔSEQco with more than 90.00% (90.19% to 92.31%) reliability at 1 year as calculated by the regression formulas (P < .001). The underestimation of measured corneal power changes was correlated with the amount of myopic correction, especially the Auto K (all P < .001).Conclusions: Direct corneal power measurements using automatic keratometry underestimated the actual corneal flattening after photorefractive surgery, which could be adjusted by a linear regression formula. Measuring the power of the first photokeratoscopic ring on videokeratography might provide a better estimation of actual corneal flattening after photorefractive surgery.