Background The two-dimensional phantom dose verification (2D-PDV) using hybrid plan and planar dose measurement has been widely used for IMRT treatment QA.Due to the lack of information about the correlations between the verification results and the anatomical structure of patients, it is inadequate in clinical evaluation.A three-dimensional anatomical dose verification (3D-ADV) method was used in this study to evaluate the IMRT/VMAT treatment delivery for nasopharyngeal carcinoma and comparison with 2D-PDV was analyzed.Methods Twenty nasopharyngeal carcinoma (NPC) patients treated with IMRT/VMAT were recruited in the study.A 2D ion-chamber array was used for the 2D-PDV in both single-gantry-angle composite (SGAC) and multi-gantry-angle composite (MGAC) verifications.Differences in the gamma pass rate between the 2 verification methods were assessed.Based on measurement of irradiation dose fluence, the 3D dose distribution was reconstructed for 3D-ADV in the above cases.The reconstructed dose homogeneity index (HI), conformity index (CI) of the planning target volume (PTV) were calculated.Gamma pass rate and deviations in the dose-volume histogram (DHV) of each PTV and organ at risk (OAR)were analyzed.Results In 2D-PDV, the gamma pass rate (3%, 3 mm) of SGAC (99.55%±0.83%) was significantly higher than that of MGAC (92.41%±7.19%).In 3D-ADV, the gamma pass rates (3%, 3 mm) were 99.75%±0.21% in global, 83.82%±16.98% to 93.71%±6.22% in the PTVs and 45.12%±32.78% to 98.08%±2.29% in the OARs.The maximum HI increment in PTVnx was 19.34%, while the maximum CI decrement in PTV1 and PTV2 were-32.45% and-6.93%, respectively.Deviations in dose volume of PTVs were all within ±5%.D2% of the brainstem, spinal cord, left/right optic nerves, and the mean doses to the left/fight parotid glands maximally increased by 3.5%, 6.03%, 31.13%/26.90% and 4.78%/4.54%,respectively.Conclusion The 2D-PDV and global gamma pass rate might be insufficient to provide an accurate assessment for the complex NPC IMRT operation.In contrast, the 3D-ADV is superior in clinic-related quality assurance offering evaluation of organ specific pass rate and dose-volume deviations.