Purpose: To evaluate the morphological changes in the cornea by in vivo laser scanning confocal microscopy (LSCM) in a large case series with silicone oil endotamponade after vitreoretinal surgery and to explore the value of LSCM in the early detection of silicone keratopathy (SK). Methods: Ninety-nine patients (99 eyes) with silicone oil endotamponade after vitreoretinal surgery were included in the current study. Slit-lamp examination and measurement of intraocular pressure (IOP) were performed first. Then the central corneas of the subjects' eyes were examined by in vivo LSCM. The analysis of images of each corneal layer was performed and the endothelial cellular density (ECD), endothelial cellular area (ECA), coefficient of variation of cell size (CoV), and percentage of hexagonal cells (PHC) were measured. Moreover, the total size of stromal deposits was measured, and the correlation between the size of deposits and the parameters of endothelial cells was analyzed. Results: Clinically recognizable abnormalities involving the cornea were identified in only 12 eyes (12.1%) under slit-lamp biomicroscopy, whereas in vivo LSCM revealed morphological abnormalities in 40 eyes (40.4%). The manifestations of endothelial lesions varied from decreased cellular density, increased polymegathism and pleomorphism to hyperreflective silicone oil membrane or droplets adhering to the endothelium. Moreover, hyperreflective deposits with various shapes could be identified in both posterior and anterior stroma, along with the infiltration of Langerhans cells beneath the epithelium. The average ECD and PHC of eyes with corneal abnormalities were significantly lower than those of normal corneas, whereas the average ECA and CoV were significantly larger (all Ps < 0.001). The patients with corneal abnormalities were significantly older than those others (P = 0.003). The rate of pseudophakic and aphakic eyes having corneal abnormalities was significantly higher than that of phakic eyes (P = 0.045). Interestingly, the total size of stromal deposits had a significant negative correlation with ECD and PHC but a significant positive correlation with ECA and CoV (all Ps < 0.001). Further correlation analysis performed in groups divided according to the lens status showed similar results and even stronger correlations in aphakic and pseudophakic eyes, whereas no statistically significant correlations were found in phakic eyes. Conclusions: In vivo LSCM was a useful tool in the early detection of corneal abnormalities caused by silicone oil injection, including varying corneal endothelium lesions and stromal abnormalities. Pseudophakic or aphakic eyes, as well as those of older patients, should receive more attention because they are more inclined to develop silicone keratopathy.