Background Using intravascular ultrasound (IVUS), we assessed the optimal stent area to prevent angiographic in-stent restenosis (ISR) after Sirolimus-eluting stent implantation for unprotected left main coronary artery (LM) disease. Method A total of 403 patients (403 LM lesions) treated with single- or two-stent strategies (crushing and T-stent) had immediate post-stenting IVUS and 9-month follow-up angiography. Ostial left anterior descending artery ([LAD], 5mm distal to carina), polygon of confluence (POC, confluence zone of LAD and left circumflex artery [LCX]) and proximal LM segment above the POC were evaluated. LCX-pullback was performed in 104 of 114 lesions with two-stent and ostial LCX (5mm distal to carina) was assessed. Post-stenting minimal stent area (MSA) was measured in each segment. Result In the overall, 46 (11.4%) showed angiographic restenosis at 9 months. Restenosis was identified in 3 (4.5%) of 67 non-bifurcation lesions with single-stent, 14 (6.3%) of 222 bifurcation lesions with single-stent cross-over, and 29 (25.4%) of 114 bifurcation lesions with two-stent. To predict ISR of the corresponding in-stent segment, the best cut-off of MSA was 6.3mm2 within the LAD ostium, 7.2mm2 within the POC, and 8.2mm2 within the proximal LM above the POC. In 104 lesions treated with two-stent technique, ISR of the LCX ostium was predicted by the MSA within the LCX ostium <5.0mm2. Using those criteria, 133 (33.8%) revealed underexpansion of at least one stented segment. Angiographic ISR (at any stented segments) was more frequent in the lesions with underexpansion vs. those without (24.1% vs. 5.4%, p<0.001). Two-year MACE-free survival rate was significantly lower in the patients with underexpansion vs. those without (89짹3% vs. 98짹1%, log-rank p<0.001). Using the multivariable Cox model, post-stenting underexpansion was an independent predictor for MACE (adjusted HR=5.56, 95% CI=1.99–15.49, p=0.001). Conclusion A smaller IVUS-MSA was responsible for the high rate of angiographic ISR and adverse cardiac events after LM stenting. With the criteria, IVUS optimization during procedure may contribute to the improvement of clinical outcomes.