Aims To define and quantify the upper lid contour by adapting Bézier curves with a newly developed software in normal subjects, assessing their reliability.

Methods Fifty eyes of 50 healthy patients with no ocular pathology were included in this study from October 2020 to November 2020. All measurements were performed on Bézier curves adjusted to the upper lid contour. An original software was used to measure the radial and vertical midpupil-to-lid margin distances (MPLD), temporal-to-nasal (T/N) ratios, contour peak location and grade of superposition (GS) and asymmetry (GA) indexes. We calculated differences in the variables measured regarding age, gender or the side of the eye being assessed.

Results The mean Bézier curve showed an excellent level of inter-rater reliability (intraclass correlation coefficient of 0.99). The median GS index of each eyelid to the mean Bézier curve was 95.4%, 8.5 IQR, and the median GA index was 3%, 3.4 IQR. The mean contour peak location was −0.35 mm, SD 0.45. Overall, the mean central MPLD was 4.1 mm, SD 0.6. No significant differences were found between male and female patients in variables derived from Bézier curves.

Conclusion Bézier curves may become a very useful tool for the assessment of upper lid contour, contour peak and symmetry. GS and GA indexes, along with the T/N area ratio are potential outcomes for this purpose. All current variables can be obtained just from one single Bézier curve measurement. Our results offer an in-depth exhaustive description of these variables and their distribution in the normal population.