PURPOSE: To compare the prediction errors (PEs) of several thick intraocular lens (IOL) formulas with a thin lens approach using variations of the same paraxial vergence calculation formula. SETTING: Department of Ophthalmology, Randers Regional Hospital, Denmark. DESIGN: Prospective, noninterventional study. METHODS: Optical low coherence reflectometry biometry was prospectively and consecutively performed in 132 eyes with subsequent phacoemulsification and insertion of the same aspheric IOL model. Clinical refraction was performed 2 months postoperatively. Retrospectively, the same paraxial vergence formula was used and only the methods were varied for calculating the postoperative IOL position, thickness, and curvatures to construct 4 formulas: Næser I formula based on thick lens calculation using the manufacturer's cutting card information, Næser II formula based on thick lens calculation using calculated IOL data from open sources, Næser III formula based on thick lens calculation and a fixed IOL thickness of 0.62 mm, and Næser IV formula based on thin lens calculation with fixed IOL position 0.31 mm anterior to the postoperative capsule. Each formula was optimized separately, hereby assuring a mean error of zero. The PE was defined as the difference between the measured and predicted spherical equivalent refraction. RESULTS: Mean absolute error amounted to 0.30 ± 0.26 diopters for all 4 formulas with no statistically significant difference. PE averaged zero for short, normal, and long eyes. CONCLUSIONS: The additional optical information provided by thick IOL calculations seems insignificant compared with other sources of error, related to the accuracy of IOL manufacturing, crystalline lens refractive index, and postoperative refraction.