The study aimed to compare the accuracy of intraocular lens (IOL) calculation to predict postoperative refraction by Optiwave Refractive Analysis (ORA), Lenstar LS 900, and the surgeon’s Modify method in normal Asian eyes. The IOL power of the Lenstar group was calculated according to Lenstar LS 900, whereas the surgeon's Modify group used topography, axial length (AL) of Lenstar, and Barrett Universal II online formula. Intraoperative aphakic measurements and IOL power calculations were obtained with the ORA system. From the results acquired through Lenstar, Modify, and ORA, the surgeon used his judgment to select the actual IOL power. Postoperative manifest refraction spherical equivalent (MRSE) was obtained 2 months after surgery. The prediction error (PE) was calculated as the difference between the postoperative MRSE and the target refraction proposed by three methods. AL, anterior chamber depth (ACD, measured from corneal endothelium to lens), lens thickness (LT), and ACD + 1/2LT were also included in the survey. In 67 eyes, the average real PE was smaller for the Lenstar (0.06 ± 0.44) and Modify (− 0.05 ± 0.40) than for the ORA group (− 0.25 ± 0.60, p < 0.05). The ORA system demonstrated the best results of IOL power selection in eyes with a normal range of ACD + 0.5 LT (5.2–5.6 mm) in Asian eyes.