Kilovoltage∞ne-beam computed tomography (kV-CBCT) based on flat-panel technology is primarily used for patient positioning. Itcan also be used for re-planning and dosime甘ic verification in adaptive radiotherapy by validating the accuracy of CBCT-image-based treatment planning. Thisstudy evalu泌的也已accuracy of dose calculation based on CBCT images using an anthropomorphic phantom. TheHounsfield unit (HU)巳lectron density calibration curves for fnll-fan and half-fan modes of CBCT a間obtained using a Catphan@ 600 phantom and compared with也已conventional CT curve. The stability of th巳CBCT calibration curve with time is also studied over a period of 8 weeks. CT and half-fan CBCT imaging of也已anthropomorphic phantom was obtained. The isodose dis甘ibutions of a single direct field and wedge fields are compared for CT- and CBCT-image-based dose calculations. Simulated identical targets and organs at risk are delineated in CT and CBCT images for which conformal radiation therapy (CRT) and intensity-modulated radiation也erapy (IMRT) plans we間generated. Thep的問tage dose differences and dose-volume histograms (DVHs) of cτ:. and CBCT-based plans are compared. Itis found that也已maximum difference in HU between half-fan CBCT and CT is 40 HD. No significant change is observed in也已calibration curves over a period of 8 weeks.立起isodose distributions computed based on CBCT and CT for a single direct field and wedge fields agrees to within 1 %. For head and neck and pelvis IMRT plans，也已dose calculated using CT agrees with that of CBCT-based calculation to within :!: 1.0%，respectively. However， for也已由or鼠，也已re is a pronounced discrepancy at也已100% isodose line and也已dose difference is within :!: 3%. Itcan be thus concluded也at CBCT images can be used for dose calculation， but it is necessary to validate the dosimetric data for inhomogeneous tissue regions .