The incidence of restenosis within stented segment after intravascular brachytherapy with recommended dose prescription protocols is up to 25%. Therefore, we designed a randomized trial comparing recommended dose prescription protocol with dosing adjusted for the source-to-target distance.Fifty-one in-stent restenosis (ISR) lesions in 48 patients underwent centered source β-irradiation with serial intravascular ultrasound. Patients randomly received 20 Gy at 1 mm either beyond lumen surface [n=25, standard group (S)] or external elastic membrane [n=26, dosing-adjusted (DA) group]. Minimum dose absorbed by 90% of adventitia (DV 90%Adv ) was calculated.DV 90%Adv was higher for the DA group than for the S group (21.63±5.67 vs. 12.05±4.88 Gy, P<.001). After 8.9±4.5 months there was complete lumen preservation in DA vs. lumen decrease subsequent to neointimal hyperplasia (NIH) in S group (0.10±1.20 vs. −0.61±1.29 mm 3 /mm, P<.05). Vessel volume increased significantly in the DA group and was unchanged in S group (+1.73, P=.002 vs. 0.14 mm 3 /mm, P=NS). DV 90%Adv correlated inversely with NIH volume and positively with vessel volume change (r=−.405, P=.007 and r=.363, P=.017, respectively).For β-irradiation of ISR, dosing adjusted for the source-to-target distance leads to significant increase in target delivered doses, which is associated with complete NIH inhibition and induction of positive vessel remodeling.