In radiation therapy it is very important to accurately measure the amount of radiation delivered. The effectiveness of the treatment depends on delivering the dose with an accuracy of 5% or better. The dosimetry in different clinics must also be consistent. For these reasons national and international calibration protocols have been developed. In the US, the American Association of Physicists in Medicine (AAPM) has published several national dosimetry protocols for the calibration of high-energy photon and electron beams. In this study the absorbed dose-to-water determined according to TG-21 and TG-51 protocols, developed by Task Group 21 and Task Group 51 of the Radiation Therapy of AAPM, are compared. The older protocol, TG-21, is based on exposure calibrated ionization chambers using a 60Co beam. Many standards laboratories have started to replace exposure standards with those involving absorbed dose-to-water. The new protocol, TG-51, is based on absorbed dose-to-water calibrated ionization chambers using a 60Co beam. Also, there are some differences between the beam quality specifiers and data proposed by the two protocols. A comparison between TG-21 and TG-51 protocols was done by determining the radiation dose rate at a designated distance for 6 and 18 MV photon beams, and 16 and 20 MeV electron beams, generated by Clinac a 2100 C linear accelerator. The cylindrical ionization chambers used in this study were Capintec PR-06G and PR-05. The results of the study show a discrepancy between the absorbed dose-to-water determined according to TG-21 and TG-51 protocols of about 1.4% and 1.7% for 6 and 18 MV photon beams, respectively. Absorbed dose-to-water determined according to TG-21 and TG-51 protocols for 16 MeV energy electron beams agree within 1.8%, while the ones of 20 MeV energy beams agree to within 2.4%. The change from exposure to absorbed dose-to-water calibrated ionization chambers has the largest impact on the differences between TG-21 and TG-51 absorbed dose-to-water, while the change in beam quality specifier and stopping power ratios have only a very small effect on these differences. The TG-51 protocol is very simple, minimizing the chance of mistakes, because it starts with absorbed dose-to-water calibration, while the TG-21 is very complex, starting with the calibration for exposure, which is different from the absorbed dose-to-water, the clinical quantity of interest. The TG-51 protocol allows the determination of a more accurate absorbed dose in a 60Co beam than the TG-21 protocol since it uses an absorbed dose-to-water calibration factor directly measured, while the exposure based dosimetry system is dependent on external physical data which are not measured in clinics.