OCT is widely used for accurate coronary stent sizing; however, the impact of coronary tortuosity or guide wire weight on the accuracy of OCT in assessing coronary dimensions is not known. This study sought to determine whether guide wire properties impact on coronary diameter and length estimation by optical coherence tomography (OCT) in tortuous segments of coronary artery. OCT coronary diameters were determined in selected patients with tortuous arteries before and after removal of a supportive guide wire to obtain unobscured images. In addition, a coronary model was created with lumen diameters of 1.7–4.8 mm. This was used to perform OCT images with a flexible (FW) and a stiff guide wire (SW) when straight or on a 25 mm radius curve to simulate tortuosity. Lumen diameter decreased by a mean of 3.2 + 1.0% after support wire removal in patients undergoing OCT for coronary sizing. In a coronary model, lumen length was constant when straight but underestimated between 0 and 4.5% (FW) and 0 and 12% (SW) when tortuous (P < 0.001, FW vs. SW, unpaired t test). Mean lumen diameter was overestimated by tortuosity (FW vs. SW for tube sizes 1.7 mm: +0–0.1 mm vs. +0.1–0.2 mm, P = 0.31; 2.7 mm: +0.02–0.17 mm vs. +0.18–0.46 mm, P < 0.01; 3.2 mm: +0.35–0.39 mm vs. +0.77–0.91 mm, P < 0.001; 4.8 mm: +0.4–0.69 mm vs. +0.88–1.05 mm, P < 0.001; unpaired t tests). Overestimation of diameter could be minimised by measuring luminal images with the least eccentric catheter position. OCT underestimation of length and overestimation of diameter should be considered when performing coronary intervention in tortuous vessels. The effect is augmented by increased wire stiffness.