Despite substantial basic and clinical efforts to address the problem of restenosis after percutaneous coronary intervention, effective preventive therapies have not yet been developed. Nevertheless, the accumulated information has provided much insight into the process of restenosis in addition to allowing standards to be developed for adequate clinical trials. The pathophysiology of restenosis increasingly appears to be distinct from that of primary atherosclerosis. Restenosis involves elastic recoil, incorporation of thrombus into the lesion and fibrocellular proliferation in varying degrees in different patients. Lack of an animal model that satisfactorily mimics restenosis is a major impediment to further understanding of the process. Clinical studies are hampered by difficulties in finding a single unifying definition of restenosis and by variable methods of reporting follow-up. Reporting of clinical outcomes of all patients in angiographic substudies would allow a more satisfactory interpretation of the results of clinical trials. Current noninvasive test results are not accurate enough to substitute for angiographic and clinical outcome data in intervention trials. In the majority of observational studies, only diabetes and unstable angina have emerged as consistently associated with restenosis; whereas most of the standard risk factors for atherosclerosis have a less consistent relation. Disappointingly, the new atherectomy and laser technologies have not affected restenosis rates. The one possible exception is coronary stenting, as a result of the larger luminal diameter achieved by the placement of the stent. In conclusion, although substantial continued effort is necessary to explore the basic aspects of cellular proliferation and mechanical alteration of atherosclerotic vessels, attention to the principles of clinical trials and observation are required to detect the impact of risk factors and interventions on the multifactorial problem of restenosis. Adequate sample sizes, collection of clinical and angiographic outcomes and factorial study designs hold promise for unraveling this important limitation of percutaneous intervention.