Background: To investigate the effect of intravitreal conbercept (IVC) injections on the incidence of postoperative vitreous hemorrhage (VH) in eyes undergoing surgery for severe proliferative diabetic retinopathy. Methods: This was a pilot prospective, comparative, and randomized study. Thirty patients, who underwent vitrectomy for severe proliferative diabetic retinopathy, were assigned randomly to either group 1 (intravitreal conbercept [IVC] injection at the end of pars plana vitrectomy) or group 2 (no IVC injection). Postoperative followup was performed on the first day, first week, first month, third month, sixth month and first year after surgery. The primary outcome was the incidence of postoperative VH. Secondary outcomes were the initial time of vitreous clearing (ITVC), best-corrected visual acuity (BCVA) and central retinal thickness (CRT) after surgery. Results: A total of 30 eyes, from 30 patients, were included. Fifteen eyes were enrolled in the IVC group and fifteen in the control group. The incidence of early and late postoperative VH was not significantly different between the control and IVC groups. ITVC was shorter in the IVC group than that in the control group, but this was not significant (7.38 ± 10.66 vs 13.23 ± 17.35, P = 0.31). Final BCVA, 1 year after surgery, showed significant improvement compared to baseline in both groups. However, analysis of the BCVA at any postoperative visit after surgery showed no significant differences between the two groups. There were two cases of recurrent VH identified at 3 and 6 months after surgery in each group, requiring a second round of surgery. Foveal thickness was significantly different between the two groups at the 3-month, 6-month and 1-year follow-up visits. Conclusions: In this pilot study, the effect of IVC injection in reducing the incidence of postoperative VH after diabetic vitrectomy at the end of vitrectomy was not shown.