Radiation delivered by intravascular stent is an appealing approach to prevent neointimal hyperplasia, since it nonselectively kills dividing cells. In particular, beta-particle-emitting radioisotope stents may prove to be an ideal means of local irradiation in that 95% of the dose is delivered within 4 mm of the stent edge and the dose drops off rapidly to < 1/1,000 of the original dose at 5 months postimplantation. In the in vitro smooth muscle cell model, one can observe a zone of growth inhibition around radioactive stent wires that averages about 6 mm at very-low-activity levels (0.006 microCi/cm of wire). In vivo studies in animal models, including porcine iliac and coronary arteries and rabbit iliac arteries, have shown the effectiveness of radioisotope stents in inhibiting neointimal proliferation. Proliferating endothelial cells appear to be relatively radioresistant. A computer model was employed to look at the radiation dose delivered as a function of distance from the stent. With very-low-activity stents, presumably, DNA of the smooth muscle cells is damaged as they migrate through the "electron fence" on the way to the neolumen, diminishing the population of myofibroblasts and reducing hyperplasia. Catheter-based radiation therapies may disable these cells before they migrate, although such an approach may not inhibit early recoil or late contraction. Based on the characteristics of beta emissions (i.e., rapid drop-off, minimal leaching), radioisotope stents containing phosphorus-32 appear to be safe. A randomized triple-blind clinical trial is planned to assess restenosis at 6 months in native coronary arteries treated with radioisotope stents.