Purpose: The accuracy of delivered dose is essential to the quality of radiotherapy treatment and tumor response. Generally, there are many types of dosimeter have been used to verify the dose from the treatment; however, most of these dosimeters are impractical for clinical situation. The goal of this study was to assess an absolute dose derived from the Monte Carlo (MC) method for the so-called 6- and 10-MV photon beams obtained from Varian Clinac 2100C linear accelerator. Methods: The deposited doses have been calculated by the EGSnrc code system and, then, were converted into the absolute doses. We were also measured, in water phantom, by an ionization chamber and, in the chest region of Rando phantom, by a thermoluminescense dosimeter (TLD). Results: The simulated data in water phantom agree with the results from both the measurement and previous studies within 2%. By comparing the absolute dose at various positions within the Rando phantom from two-opposing irradiated fields, the difference from MC calculation and TLD measurement was within 2%. Unfortunately, the calculated doses obtained from the collapse cone convolution (CCC) algorithm showed notable difference from that of the MC method. For the interface region within the provided field, it was higher than that from the MC method by almost 5% for the 6-MV and 7% for the 10-MV photon beam. Conclusion: Our findings indicated that the MC method was on the level with the measurement for the dose determination, especially within the delivered field to a heterogeneous phantom.