Purpose: To evaluate the possible dosimetric advantage of volumetric modulated arc therapy (VMAT) over conventional 3D conformal therapy for left breast tumor bed boost treatment. Methods: Ten patients were retrospectively planned with VMAT technology for treatment of left breast after lumpectomy. 6 patients were simulated under deep inspiration breath hold, 4 patients were free breath. In each plan, a partial ARC of 200° to 220° was used. All VMAT plans were normalized to deliver 100% of the 16Gy in 8 fractions prescription dose to 95% of PTV volume. The plans were evaluated by calculation of heart maximum dose (Dmax_heart), percentage of lung volume receiving 10% of prescribed dose (V10_lung), percentage of ipsilateral breast receiving 50% of prescribed dose (V50_breast), maximum dose (Dmax), and the location of hot spot. These dosimetric parameters were compared with those for clinical plans using conventional 3D conformal technique. Results: For these ten patients (mean breast volume 1198 cc, mean boost PTV 161 cc), the mean dosimetric parameters for ARC plans/3D plans were 8.8%/28.5% of prescription dose for Dmax_heart, 5.4%/17.7% of total lung volume for V10_lung, 35.8%/40.0% of ipsilateral breast volume for V50_breast, 111.2%/111.6% of prescription dose for Dmax. Hot spots were located inside PTV in all 10 cases for ARC plans, but only in 6 out of 10 cases for 3D conformal plans. Conclusion: Initial results suggest that partial arc VMAT plans have dosimetric advantages for breast boost treatment. The plan quality was superior to conventional 3D conformal plans. In addition, with a template of standardized ARC beam arrangement and dose constraints for optimization, treatment planning and delivery for breast tumor bed boost can be more efficient.