PurposeTo assess the repeatability, reproducibility, and agreement of central corneal thickness (CCT) measured by non-contact Fourier domain optical coherence tomography (FD-OCT; OptoVue) with the other two contact devices, ocular response analyzer (ORA; Reichert Ophthalmic Instruments) and Ultrasound Pachymetry (USP; DGH Technologies). MethodsThis observational cross-sectional study measured CCT sequentially using FD-OCT, ORA and USP. The first 16 volunteers (32 eyes) received three measurements by two independent examiners in a single session to determine intra-observer repeatability and inter-observer reproducibility. An additional 27 volunteers (54 eyes) received one measurement by the same examiner. The measurements of all 86 eyes were analyzed for the difference, correlation, and agreement among the three devices. ResultsFD-OCT measured the thinnest while USP measured the thickest CCT (548.6 ± 28.3 μm, 556.9 ± 28.8 μm, and 560.0 ± 28.8 μm by FD-OCT, ORA, and USP, respectively, p < 0.001). The mean differences (lower/upper limit of agreement) for CCT measurements were 8.4 ± 7.6 μm (−6.5/23.2) between ORA and FD-OCT, 11.4 ± 7.3 μm (−2.8/25.7) between USP and FD-OCT, and 3.1 ± 5.1 μm (−6.9/13.1) between ORA and USP. The intra-class correlation coefficients were above 0.98 for all tested groups. FD-OCT had the lowest intra-examiner variability (coefficient of repeatability of 0.64%) and lowest inter-examiner variability (coefficient of reproducibility of 1.16%). ConclusionFD-OCT, ORA, and USP demonstrated good inter-observer reproducibility and intra-observer repeatability. The three measurements were highly correlated; however, systematic differences between the three tested devices did exist. FD-OCT was a reliable and examiner-independent method in CCT measurement.