Background Early lumen loss (ELL) may be detected in patients undergoing coronary interventions for in-stent restenosis. This phenomenon may contribute to recurrences. This prospective, randomized study was designed to determine if repeat stent implantation may prevent ELL in patients with in-stent restenosis. Methods Forty patients were randomized: 20 were allocated to elective stent implantation and 20 to conventional balloon angioplasty. Quantitative coronary angiography and intravascular ultrasound (IVUS) volumetric studies were systematically performed (1) before the procedure, (2) immediately after intervention, and (3) 30 to 60 minutes-later. Results Baseline characteristics were similar in both groups. After the delay time interval (46 8 minutes), quantitative coronary angiography revealed a significant reduction in minimal lumen diameter (2.2 +/- 0.5 mm vs 1.7 +/- 0.5 mm, P <.001) in the balloon angioplasty arm. Likewise, in this group, IVUS demonstrated a reduction in mean lumen area (7.1 +/- 2 mm vs 6.2 +/- 2 min, P <.001) and lumen volume (144 59 mm vs 126 54 min, P<.001). In 4 of these patients, ELL was severe enough to require further intervention. In the stent arm, however, angiographic data and IVUS mean lumen area (7.7 +/- 3 mm(2) vs 7.7 +/- 3 mm(2)) and lumen volume (161 +/- 72 mm(3) vs 160 +/- 69 mm(3)) remained unchanged after the delay time interval. On multivariate analysis, stent implantation was an independent predictor of the absence of ELL by quantitative coronary angiography and by IVUS. In addition, patients with a larger ELL on IVUS had a lower event-free survival at 1 year (40% vs 79%, log rank P =.003). Conclusions This randomized study demonstrates that (1) ELL is frequently detected after treatment of in-stent restenosis with balloon angioplasty, that (2) ELL influences the long-term clinical outcome of these patients, and that (3) repeat stent implantation prevents ELL.