Purpose To compare the refraction before and after cycloplegia with 0.5% and 1.0% cyclopentolate eye drops using five different measurement modalities. Methods This prospective, clinical comparative study enrolled 96 eyes of 48 healthy patients with a mean age of 26.6 +/- 4.21 years (range: 19-34). Subjective refraction, retinoscopy, and objective refraction were measured using three autorefractometers: Topcon KR-800 (TC), Retinomax K-plus 3 (RM + 3), and Retinomax K-plus Screeen (RM + S) under noncycloplegic and cycloplegic conditions. Cycloplegia was performed in the right eye using 0.5% and in the left eye with 1.0% cyclopentolate eye drops. Differences in refraction in noncycloplegia and cycloplegia, between cycloplegia with 0.5% and 1.0% cyclopentolate, and between the devices were investigated. Results Cycloplegic mean spherical equivalent was -1.77 +/- 2.34 diopters (D) (-9.75 to + 1.625). All approaches showed a statistically significant hyperopic shift (p < 0.001, each) after induction of cycloplegia using both regimes. Lowest median (interquartile range) hyperopic shift was shown by TC (0.25 D (0.38)) and retinoscopy (0.25D (0.75)), and the highest by RM + 3 (0.75 (1.31)). No statistically significant differences between cycloplegia using 0.5% and 1.0% regimens were shown in all modalities (p > 0.05, each). In noncycloplegia, there were greater differences compared to cycloplegia. No influence of iris color on the refraction was found. Conclusion After induction of cycloplegia all devices showed a hyperopic shift and good comparability to retinoscopy. In all measurement modalities, no significant refraction differences between 0.5% and 1.0% cyclopentolate eye drops were seen. Therefore, 0.5% cyclopentolate was proven to have a sufficient effect with presumably better tolerability.