We used intravascular ultrasound (IVUS) to compare the expansion of drug-eluting stents (DES) implanted to treat bare-metal stent (BMS) restenosis after 3 common re-stenting strategies. A total of 80 consecutive BMS restenotic targets were re-stented either directly (n = 30, group 1) or after low-pressure (<8 atm) pre-dilation with an undersize regular balloon (n = 16, group 2) or after high-pressure (>12 atm) pre-dilation with a semicompliant balloon the same or greater diameter as the original BMS diameter (n = 34, group 3). More patients from groups 2 and 3 had diabetes mellitus. The targets in group 1 were more proximal and focal. Lesions from groups 2 and 3 were more severe. The size and deployment pressure of the DESs and the achieved angiographic results were all similar. The post-intervention minimum stent area and the percentage of expansion of DES (minimum stent area/distal reference lumen area x 100%) were, however, both significantly larger in group 3 (6.4 +/- 1.5 mm(2) in group 3 vs 5.6 +/- 1.6 mm(2) in group 1 vs 4.4 +/- 1.4 mm(2) in group 2, p <0.001; and 88 +/- 30% in group 3 vs 74 +/- 14% in group 1 vs 73 +/- 23% in group 2, p = 0.021). A post-intervention minimum stent area <5.0 mm(2) was seen in only 3 lesions in group 3 (8.8%) versus 14 in group 1 (46.7%) and 11 in group 2 (68.8%; adjusted odds ratio 0.11, 95% confidence interval 0.03 to 0.38, p <0.001). Overall, the acute lumen gain was mainly from BMS re-expansion; however, the BMS volume increased the most in group 3 (p <0.001). In conclusion, high-pressure pre-dilation leads to superior post-intervention expansion of DESs implanted to treat BMS restenosis, regardless of the original expansion because of the greater BMS re-expansion.