Résumé : Dry eye syndrome is a highly prevalent disease of the ocular surface characterized by an instability of the tear film. Traditional methods used for the evaluation of tear film stability are invasive or show limited repeatability. Here we propose a new non-invasive fully automated approach to measure tear film thickness based on spectral domain optical coherence tomography and on an efficient delay estimator. Silicon wafer phantom were used to validate the thickness measurement. The technique was applied in vivo in healthy subjects. Series of tear film thickness maps were generated, allowing for the visualization of tear film dynamics. Our results show that the in vivo central tear film thickness measurements are precise and repeatable with a coefficient of variation of about 0.65% and that repeatable tear film dynamics can be observed. The presented approach could be used in clinical setting to study patients with dry eye disease and monitor their treatments.