Optical surface monitoring system (OSMS) could achieve continuous monitoring of patient position through a nonradiographic and noninvasive technology, which has been widely applied to many sites of radiation therapy including breast, intracranial, head and neck, abdomen, and the extremities. However, no study has investigated the application of OSMS in radiation therapy of vulvar cancer. In this study, we hypothesized using OSMS would reduce setup error in the vulvar cancer compared to standard laser-based setup (SLS). A single-center prospective trial included 10 patients with vulvar cancer treated with intensity modulated radiation therapy (IMRT) for a total of 250 fractions. Every patient was immobilized with Styrofoam in a frog-leg position and used marks on the patient's skin to align to the room lasers for pre-setup. The treatment couch values of lateral (Lat), longitudinal (Lng), vertical (Vrt) in three translation directions and Rtn (the rotation along the z axis), Pitch (the rotation along the x axis), Roll (the rotation along the y axis) in three rotations directions were recorded. Then completed the setup of patient with information from a surface imaging system and obtained deviation values. Setup is followed by pretreatment cone-beam computed tomography (CBCT) to further verify patient and target position by registering to the planning CT with respect to bony anatomy. With CBCT as a gold standard for patient setup, the absolute values of setup error and error distribution of SLS method and OSMS method from the same patient in above six directions were compared. The mean of absolute values of setup errors in the SLS method were significantly higher than those of OSMS method in Lat (0.34cm vs 0.20cm), Lng (0.42cm vs 0.29cm), Vrt (0.43cm vs 0.28cm) and Rtn (0.90° vs 0.67°) directions (P<0.05). The mean errors in SLS and OSMS methods in Pitch and Roll directions were 0.93° and 1.02° (P = 0.16), 0.42° and 0.45° (P = 0.48), respectively, with no significant difference. In the OSMS method, the distribution rates of the setup errors within 0.5cm in three translation directions and within 1.5° in three rotations directions were significantly higher in Lat (90.8% vs 80.4%), Lng (82.0% vs 68.8%), Vrt (83.2% vs 65.6%), and Rtn (91.2% vs 80.4%) directions than those of SLS method (P<0.05). In the Pitch (79.2% vs 80.4%) and Roll (97.2% vs 98.0%) directions, the distribution rates of the OSMS method were slightly lower than those of the SLS method, with no significant difference (P = 0.738 and P = 0.559). The implementation of OSMS in IMRT for vulvar cancer could reduce setup error and increase precision of setup, especially for the directions of Lat, Lng, Vrt and Rtn.