To achieve a good clinical outcome in radiotherapy treatment, a certain accuracy in the dose delivered to the patient is required. Therefore, it is necessary to keep the uncertainty in each of the steps of the process inside some acceptable values, which implies a global uncertainty as low as possible. This work is focused on the uncertainty evaluation of absorbed dose to water in the routine calibration for clinical beams, in the range of energies used in external radiotherapy. With this aim, different uncertainty components (corrected electrometer reading, calibration factor, beam quality correction factor and reference conditions) associated to beam calibration have been considered. Results show a typical uncertainty in the determination of absorbed dose to water during beam calibration around 1.3% for photon beams and 1.5% for electron beams (k=1 in both cases) when the N(D,w) formalism is used and is theoretically calculated. These values may be different depending on the uncertainty provided by the standards laboratory for calibration factor, which is shown in the work. If the total application of the N(D,w) formalism, that is to say, specific calibrations of each chamber in the user's beam qualities, is taken into account the uncertainty in this step of the process could be placed close to 1.0%. Furthermore, the possibility of an uncertainty reduction with the absorbed dose to water formalism adoption against the air kerma one is discussed.