A large matched‐cohort study was carried out to determine correlates of in‐hospital abrupt vascular closure (AC). Univariate analysis identified current cigarette smoking (P = 0.021), myocardial infarction within 24 hr prior to procedure (P = 0.0035), emergency procedure (P = 0.02), lesion thrombus (P = 0.0001), and lesion angulation (P = 0.021) as significant clinical and angiographic variables. Relative to balloon angioplasty (PTCA), use of atherectomy (P = 0.015) and laser devices (P = 0.018) but not elective stent placement (P = 0.97) were associated with increased risk of AC. In the multivariate model, current cigarette smoking (P = 0.0474), lesion thrombus (P = 0.0001), lesion angulation (P = 0.0124), use of atherectomy devices (P = 0.001), and laser devices (P = 0.0037) remained as significant correlates of increased AC events. In conclusion, the risk of AC appears associated primarily with lesion characteristics and use of nonballoon devices other than stents. Elective stent placement did not appear to reduce AC risk over conventional PTCA; the small number of patients studied may have prevented any benefit from being observed. Cathet. Cardiovasc. Intervent. 47:391–395, 1999. © 1999 Wiley‐Liss, Inc.