Abstract Background Percutaneous coronary intervention (PCI) with a drug-eluting stent (DES) is increasingly being used for the treatment of coronary artery diseases. However, the role and effectiveness of PCI with a bare metal stent (BMS) have not yet been established. Methods Among the 2197 patients (2653 lesions) treated with PCI from 2003 to 2012 at three institutions in Fukuoka, 859 patients (1032 lesions) without acute coronary syndrome and in whom we were able to perform follow-up coronary angiography after stent placement and collect detailed data were selected for this study. Among the patients treated by BMS, the cut-off lesion reference for the development of in-stent restenosis (ISR) was calculated by comparing patients with and without ISR. In all patients with a lesion reference above/below the cut-off point of BMS-ISR, medium-term clinical and angiographic outcomes were compared between the BMS and DES groups. Results In patients treated with a BMS, the lesion reference odds ratio 0.68, 95% confidence interval (CI) 0.43–0.97, p = 0.03 was strongly correlated with ISR, and the area under the receiver operating characteristic curve was 0.704, with moderate accuracy. Calculation of the cut-off lesion reference for BMS-ISR, using a relative cumulative frequency distribution and a sensitivity/specificity curve, showed that the cut-off lesion reference for BMS-ISR was 3.08 mm, and the sensitivity/specificity was 60.4%. For the DES and BMS groups in cases with a lesion reference over 3.08 mm, BMS placement did not correlate with ISR (odds ratio 0.98, 95% CI 0.85–1.12), and there was also no correlation between BMS and major adverse cardiac events. Conclusion In patients with a lesion reference ≧ 3.08 mm, medium-term clinical and angiographic outcomes were equivalent between the use of a DES and BMS. Thus, BMS placement is well worth considering, especially in patients with a lesion reference ≧ 3.08 mm, in whom double-antiplatelet therapy needs to be stopped at an early stage.