The aim of this study was to perform a comparison between robust optimization and planning target volume (PTV)-based optimization plans using volumetric modulated arc-therapy (VMAT) by evaluating perturbed doses induced by localization offsets for setup uncertainties in larynx cancer radiation therapy. Methods: Ten patients with early-stage (T1-2N0) glottis carcinoma were selected. The clinical target volume (CTV), carotid arteries, and spinal cord were contoured by a radiation oncologist. PTV-based and robust optimization plans were normalized at D 50% to the PTV and D 98% to the CTV, respectively. Both optimization plans were evaluated using perturbed doses by specifying user defined shifted values from the isocenter. CTV dose (D 98% , D 50% , and D 2%), homogeneity index (HI) and conformity index (CI 95% , CI 80% , and CI 50%), as well as doses to the carotid arteries and spinal cord were compared between PTV-based and robust optimization plans. Results: The robust optimization plans exhibited superior CTV coverage and a reduced dose to the carotid arteries compared to the PTV-based optimization plans (p < 0.05). HI, CI 95% and the dose to the spinal cord did not significantly differ between the PTV-based and robust optimization plans (p > 0.05). The robust optimization plans showed better CI 80% and CI 50% compared to the PTV-based optimization plans (p < 0.05). Plan perturbed evaluations showed that the robust optimization plan has small variations in the doses to the CTV, carotid arteries, and spinal cord compared to the PTV-based optimization plan. Conclusions: The robust optimization plan may be a suitable treatment method in radiotherapy for larynx cancer patient.