OBJECTIVE: To investigate whether an elective change in the anticoagulation protocol for patients with coronary stents affected clinical outcomes and length of hospital stay. DESIGN: Retrospective observational study of a consecutive series of patients treated with coronary stents over an 18 month period from April 1994 to October 1995. BACKGROUND: Intensive anticoagulation regimens are used in many UK centres to reduce the risk of coronary stent thrombosis. Recent data have called into question the necessity for full anticoagulation and favourable results have been reported with antiplatelet agents alone. The results from a tertiary referral centre were investigated during a period where an elective change in policy was made: an initial 70 patients were treated intensively with intravenous heparin and with warfarin and aspirin; subsequently 94 were treated with aspirin and deployment of a high pressure balloon only. METHODS: Review of case notes, angiograms, and a database of intervention procedures and telephone interview. Classic epidemiological techniques, as well as linear regression and logistic regression, were used to model the outcomes of major procedural complications and length of hospital stay. PATIENTS: 164 patients treated with 196 coronary stents. RESULTS: There were 22 (13.4%) major complications (coronary bypass grafting 11, subacute thrombosis 6, tamponade 2, myocardial infarction 1, death 2). With logistic regression, the risk of major complication was shown not to be affected by anticoagulation (relative risk (RR) 1.03; P = 0.97). Significant determinants of risk included acute vessel closure as an indication for stenting (RR = 80.6; P < 0.001) and sex (male: female RR = 0.19; P = 0.02). The median length of stay (LOS) was 5 days (1-45). Use of a linear regression model showed that anticoagulation added 4.5 days and a major complication added a further 4.5 days to a baseline length of stay of 3.2 days (R2 = 0.32; P < 0.001). CONCLUSION: This is a report of coronary stenting as part of usual clinical practice in one British tertiary referral centre. In this experience, treatment with aspirin alone is probably as safe as intensive anticoagulation, and has the benefit of reducing length of stay by more than 50% to 3.2 days in an uncomplicated case.