Frequency-domain optical coherence tomography (FD-OCT) is a novel technology which provides high-resolution cross-sectional images of coronary arteries. The aim of this study was to evaluate the inter-scan reproducibility of geometric FD-OCT measurements in the clinical setting. We examined 20 coronary lesions using FD-OCT. Following the FD-OCT image acquisition (1st pullback), and after the disengagement and re-engagement of the guiding catheter, an additional acquisition (2nd pullback) was performed using a new FD-OCT catheter. There was excellent correlation for minimum lumen area (r = 0.99, P < 0.001), lesion length (r = 0.99, P < 0.001) and lumen volume (r = 0.99, P < 0.001) between the 1st pullback and the 2nd pullback. The Bland-Altman test demonstrated good agreement between the 1st pullback and the 2nd pullback: the mean difference for minimum lumen area, lesion length, and lumen volume was 0.05 mm(2), 0.03 mm, and 0.70 mm(3), respectively; and the lower and upper limit of agreement for minimum lumen area, lesion length, and lumen volume was -0.58 and 0.48, -0.36 and 0.42, and -13.4 and 12.1, respectively. FD-OCT showed an excellent inter-scan reproducibility for the geometric coronary artery measurements. Our findings emphasize the value of FD-OCT as a tool for clinical longitudinal studies of coronary artery disease.