This study investigated the inherent properties of pupillary red reflexes captured by smartphone cameras for refractive error detection. This cross-sectional study was conducted at the Public Eye Hospital in Pekanbaru, Indonesia. Red reflex images were captured using an iPhone 6s with the same protocol (in a dark room, in portrait orientation, and without cycloplegic agents). A total of 651 images were acquired from 350 patients. High myopia (88.2% images) predominantly appeared on the right side of the pupil, whereas low and high hyperopia (91.8% and 92.9% images, respectively) predominantly appeared on the left side of the pupil. The red reflex fills the entire pupil in low myopia and emmetropia (75.4% and 84.5% images, respectively). The left red reflex significantly (p < 0.001) has a positive diopter power (+1.361 ± 1.925 D), while the total and right red reflexes have a negative one (-1.174 ± 1.325 and-5.977 ± 3.655 D, respectively). The total red reflex significantly (p < 0.001) has a wider diameter ratio (0.617 ± 0.081) compared with the right and left red reflex (0.246 ± 0.066 and 0.299 ± 0.087). Spearman’s correlation analysis showed a moderate correlation between the crescent size and dioptric power (a correlation coefficient of 0.615 and an R2 value of 0.314 (p < 0.001) in the right and total red reflex; a correlation coefficient of-0.504 and an R2 value of 0.011 (p < 0.001) in the left and total red reflex). Different red reflex patterns can quickly identify refractive errors. © 2025 Optica Publishing Group.