Objective  To assess the role of an oculus keratograph in the measurement of pupil size under scotopic illumination and pupil reflection at 3 photostimulation modes. Methods  One hundred twenty-five subjects （125 eyes） were enrolled in this cross-sectional study． Pupil parameters were measured by two ophthalmologists under 3 photosti-mulation modes （scotopic conditions， pupillogram and anisocoria）. An independent sample t test and ANOVA test were used to verify the differences in pupil parameters between genders and refractive error groups. The differences in pupil size after different stimulations were analyzed by a paired t test. The reproducibility of the measurements was analyzed by an intraclass correlation coefficient （ICC） and Bland-Altman plot. Results  The differences were not statistically significant for the largest pupils under the 3 conditions （F=1.03， P＞0.05）. Under scotopic conditions， the minimum， maximum and range of pupil size were 5.34±0.95 mm， 6.18±0.85 mm and 0.86±0.37 mm， respectively. Under pupillogram mode， the pupil diameter of minimum and maximum， and range of pupil size and pupil diameter 1 second after stimulation were 3.57±0.74 mm， 6.10±0.92 mm， 2.54±0.40 mm and 2.34±0.46 mm， respectively. For anisocoric conditions， the minimum， maximum and range of pupil size and pupil diameter for 1 and 5 seconds after stimulation were 2.69±0.45 mm， 5.94±0.84 mm， 3.27±0.55 mm， 2.72±0.58 mm and 2.94±0.47 mm， respectively. There was a statistically significant difference in pupil size for 1 and 5 seconds with anisocoria （t=0.38， P＜0.01）. The ICCs of intraobserver and interobserver were all higher than 0.800. Pupil size was not correlated with gender or refractive error. Conclusion  The oculus keratograph can be used to measure pupil size and pupil reflection under different photostimulations and illumination levels with a high degree of reproducibility．