Abstract This paper describes a method for determining the energy of the X-ray spectrum end-point (E end) by means of X-ray spectrometry performed with a CdTe detector. Designated the spectrometric practical end-point (SPE) method, this technique is based on the comparison of the high-energy part of the measured detector spectrum with a detector spectrum obtained using Monte Carlo calculations. It combines findings and results from a detector characterization previously done for the purpose of unfolding the X-ray fluence spectra. Here, the influences of tube current, tube voltage ripple, detector count rate, and the choice of calibration radionuclide sources were studied. The findings for the latter two parameters led to the modification of a procedure for energy calibration of the detector as well as to updated parameters of the charge carrier transport (CCT) and Gaussian energy broadening (GEB) models valid for high-count-rate measurements. Additionally, the influence of improper energy calibration and of neglecting the CCT and GEB effects on the main characteristics of the unfolded X-ray fluence spectra was analyzed to provide an overview of systematic errors. A quantity obtained with the SPE method, the spectrometric practical voltage (SPV), is introduced. It is demonstrated that the SPV agrees with the practical peak voltage within 0.03% when the voltage ripple of a high-voltage generator is below 2%. A comparison done at PTB, Germany, of the SPV to voltage divider readings showed relative and absolute differences between the reading and the SPV of between 0.1% and 0.8% (0.15 keV to 1.16 keV) in the 20 to 300 kV tube voltage range. Its combined uncertainty varied between 0.13 and 0.33 keV for continuous beam N-series qualities ranging from 10 to 300 kV. The SPE method was then used to determine E end of N-series qualities realized at Czech Metrology Institute. Also, the difference between E end as obtained with the SPE method and that found with a simple linear fit of the straight high-energy part of the detector spectra was evaluated, revealing a difference of between -0.2 and +1.5 keV for N-series qualities up to 300 kV. The SPE method is suitable for users who have already implemented an in-beam X-ray spectrometry technique and seek to determine the tube voltage but do not possess a voltage divider.