Intravascular brachytherapy is the primary treatment for coronary in-stent restenosis. Variations in dose in the treated artery may represent a potential cause of treatment failure. We compared dose distributions in patients who had developed recurrent restenosis (treatment failure) with those in patients who remained event free at 9 months (treatment success). We followed 140 patients who were receiving brachytherapy for in-stent restenosis with 4 radiation delivery devices to identify treatment failures and successes. Through a nested case-control construct, treatment failures (n = 14) were compared 1:2 with treatment successes (n = 28) matched by radiation delivery system and in-stent restenosis lesion pattern. The dose absorbed by 90% of the artery encompassed by the external elastic membrane (D(90)EEM) was calculated by applying intravascular ultrasound at 2-mm intervals along the treated lesion. Dose calculations were performed using dose kernel integration techniques generated from Monte Carlo simulations. The mean minimum D(90)EEM in treatment failures was 7.46 +/- 1.98 Gy, and that in treatment successes was 8.87 +/- 1.13 Gy (p = 0.007). Using a minimum dose threshold of 8.4 Gy, a minimum D(90)EEM <or=8.4 Gy occurred in 13 patients (93%) whose treatment failed but only in 9 patients (32%) whose treatment was a success (p <or=0.001). No confounding variables were found to be statistically significant between treatment failures and successes. In conclusion, current brachytherapy dose prescriptions allow for inter- and intralesion variations in dose. Arteries that receive <or=8.4 Gy at any point along the external elastic membrane are more likely to result in treatment failures. Dosimetry guided by intravascular ultrasound may be critical to ensure adequate dose delivery and outcomes.