To evaluate differences in the bacterial contamination rates of the vitreous cavity between patients undergoing transconjunctival 25-gauge microincision vitrectomy surgery (MIVS) and conventional 20-gauge pars plana vitrectomy (PPV). Prospective, comparative, consecutive, interventional case series. Eighty-one eyes of 81 patients who underwent primary vitrectomy and completed perioperative sample collection. Patients were randomly assigned to 25-gauge MIVS or 20-gauge PPV. Conjunctival swabs were obtained from each patient before and after preoperative administration of topical 0.5% moxifloxacin. Vitreous samples were collected at the beginning and end of surgery. All 4 consecutive specimens from each eye were cultured using direct culturing techniques under aerobic and anaerobic conditions. The primary outcome measure was the incidence of bacterial contamination of the vitreous cavity at the start and end of vitrectomy. The secondary measures were the incidence of bacterial contamination of the ocular surface and the disinfection rate with preoperative moxifloxacin. Of the 81 eyes (40 eyes in the 25-gauge MIVS group; 41 eyes in the 20-gauge PPV group), the incidences of positive bacterial isolation at the 4 time points of sample collection were 77.5%, 62.3%, 22.5%, and 0% in the former group and 82.9%, 63.4%, 2.4%, and 0% in the latter group. Although the rate of bacterial contamination of the ocular surface significantly (P<0.001) decreased after preoperative moxifloxacin administration in both groups, transconjunctival 25-gauge MIVS had a significantly (P = 0.007) higher incidence of vitreous contamination at the beginning of surgery compared with conventional 20-gauge PPV. The multivariate model showed that 25-gauge MIVS (odds ratio, 11.27; P = 0.027; 95% confidence interval, 1.31-96.79) was the factor prognostic of vitreous contamination at the beginning of surgery. Propionibacterium acnes was identified most often in the vitreous samples (80% of cases), which was consistent with the commensal bacteria isolated from the ocular surface. The higher incidence of bacterial contamination of the vitreous cavity at the beginning of 25-gauge MIVS suggests the increasing risk of direct inoculation of ocular surface flora into the vitreous cavity through the transconjunctival trocar-cannula system compared with conventional 20-gauge PPV. However, vitreous cavity contamination can be eliminated during vitrectomy in most cases. Proprietary or commercial disclosure may be found after the references.