AIM: To identify the predictive factors and laser photocoagulation associated with the use of silicone oil as endotamponade during primary diabetic vitrectomy. METHODS: The medical and surgical records of 690 patients (798 eyes) who underwent primary diabetic vitrectomy at a tertiary eye hospital in China from January 2018 to December 2018 were reviewed in this retrospective cohort study. The patients’ baseline characteristics and preoperative treatments were recorded. The binary Logistic regression model was used to evaluate the risk factors for the use of silicone oil as endotamponade agent during primary vitrectomy for proliferative diabetic retinopathy (PDR)-related complications. RESULTS: Among 690 patients with mean age of 52.1±10.5y (range: 18-85y), 299/690 (43.3%) were female. The 31.6% of the eyes received preoperative laser treatment, and 72.4% of the eyes received preoperative anti-VEGF adjuvant therapy. Non-clearing vitreous haemorrhage (VH) alone or combined with retinal detachment was the main surgical indication (89.5%) for primary vitrectomy. Silicone oil was used as endotamponade in 313 (39.2%) eyes. Lack of preoperative laser treatment [odds ratio (OR) 0.66, 95% confidence interval (CI): 0.48-0.92; P=0.015] and older age (OR 0.96, 95%CI: 0.95-0.98; P<0.001) were predictors of silicone oil tamponade during primary vitrectomy for PDR. CONCLUSION: The lack of preoperative laser treatment is a significant predictor of silicone oil tamponade during primary vitrectomy for PDR. However, the severity of PDR relevant to silicone oil use should be further evaluated.