Objective:To analyse the neointimal coverage of sirolimus-eluting stent (SES) and bare-metal stent (BMS) visualised in vivo by optical coherence tomography (OCT).Methods:OCT images were obtained in 26 coronary vessels of 24 patients at 5–93 months after SES or BMS deployment. The short-term BMS group (BMS1) consisted of eight BMS in seven patients at 5–10 months of follow-up, the long-term BMS group (BMS2) consisted of six BMS in six patients at 23–93 months of follow-up, and the SES group (SES) consisted of 13 SES in 10 patients at 6–12 months of follow-up. The strut apposition, strut coverage and mean maximal and minimal neointimal thicknesses (NIT) for both BMS groups and SES were compared.Results:OCT images were acquired successfully. Significant differences between completely apposed and malapposed stent struts (p<0.0001) and between covered and uncovered stent struts (p<0.0001) were found among the three groups. The mean maximal and minimal NIT in the SES group were all significantly less than those of the BMS1 or BMS2 group, the minimal NIT in the BMS1 group was significantly less than that of the BMS2 but the mean maximal NIT was no significant difference between the BMS1 and BMS2 groups. In an open bifurcation artery, 19 struts visualised by OCT had no discernible coverage or were surrounded by either thrombus or a thick tissue layer.Conclusions:OCT imaging can clearly visualise stent apposition and neointimal coverage of stent struts. Incomplete strut apposition and lack of strut coverage occurred with a significantly higher frequency in SES than in BMS. These findings may explain the occurrence of late thrombosis in SES.