Stents may be undersized during primary percutaneous coronary intervention (PCI) in ST-segment elevation myocardial infarction (STEMI), leading to higher rates of stent thrombosis. We sought to compare the reference vessel diameter (RVD) of the infarct-related artery before, immediately after, and at late follow-up after PCI for STEMI. We further investigated whether vessels treated with paclitaxel-eluting (PES) or bare-metal stents (BMS) behave differently with respect to RVD at follow-up. From the HORIZONS-AMI trial, we identified 2,974 patients (3,589 lesions) with complete quantitative angiographic data for stent implantation (2,233 treated with PES and 741 treated with BMS). Considering all lesions, the median RVD was 2.87 mm (25th-75th percentile 2.54-3.22 mm) at baseline, 2.92 mm (2.58-3.28 mm) immediately post-PCI, and 2.88 mm (2.55-3.22 mm) after 13 months (1,197 patients; P = .001 pre vs post, P = .06 post vs follow-up, and P = .21 pre vs follow-up). There were no significant differences between the RVD for PES versus BMS at any period. The maximal stent or balloon size was 3.00 mm (3.00-3.50 mm) for both groups. There were no differences in RVD at baseline or post-PCI between patients with and without stent thrombosis. Reference vessel diameter does not change substantially from baseline to follow-up, irrespective of stent type. Stent diameter was appropriate for vessel size. The RVD of patients with and without stent thrombosis was similar at baseline and post-PCI. Thus, the high rates of stent thrombosis after primary PCI for STEMI cannot be attributed to stent undersizing.