PURPOSE: To evaluate the agreement of automated refraction by the Chronos device (Topcon Corporation, Tokyo, Japan) with standard refraction techniques. METHODS: This retrospective, single-center study recruited patients who underwent refraction with the Nidek ARK-1 autorefraction [Nidek Technologies, Gamagori, Japan] and manual subjective refraction (standard method group), followed by autorefraction and guided subjective refraction with the Chronos binocular refraction system [Topcon Corporation, Tokyo, Japan] (Chronos method group). Concordance rates were set based on mean absolute differences of refractive measurements. Bland-Altman analysis, non-parametric analysis and Pearson's correlation coefficient (r) were used to assess agreement between methods. RESULTS: In all, 361 eyes of 182 patients (mean age 37.4±17.5 years) were included. Slightly more myopic mean objective and subjective sphere (P=0.005
P<
0.001), cylinder (P=0.001
P<
0.001) and spherical equivalent (SE
P<
0.001 for both) values were observed with the Chronos method. No statistically significant differences were seen for objective and subjective J0 (P=0.947
P=0.312) and objective and subjective J45 (P=0.131
P=0.084) values. High concordance between methods (difference ≤0.25 D) was greater for objective sphere (79.0%), cylinder (87.1%), and SE (71.7%), than for subjective measurements (65.7%, 80.8%, and 57.0%, respectively). The methods were strongly correlated for objective and subjective sphere (r=0.983, P<
0.001
r=0.971, P<
0.001, respectively) and for objective and subjective cylinder (r=0.936, P<
0.001
r=0.869, P<
0.001, respectively). Good agreement was observed between the two methods with low systematic bias and clinically acceptable mean differences. CONCLUSION: The results showed good agreement and strong correlations between both methods, suggesting that the Chronos device can be a good alternative technique for refractive error screening in routine clinical practice.