The M spectrum of the element uranium was reinvestigated by using both high-resolution wavelength dispersive (WD) spectrometry as well as energy dispersive (ED) spectrometry. Thereby we observed relative intensities that deviate from data in the literature. These discrepancies were not only observed for the weak U M lines but also for major lines. By measuring the Mα,β region of the spectrum with a PET crystal in second-order reflection, a sufficient energy resolution was achieved to separate Mα2 (M5N6) from Mα1 (M5N7). The intensity ratio I(M5N6)/I(M5N7) was determined to be approximately 5%, which is in strong contrast to the data tabulated by White and Johnson [White, E.W. & Johnson, G.G. (1970). X-Ray and Absorption Wavelengths and Two-Theta Tables. ASTM Data Series DS37A, 2nd ed. Philadelphia, PA: American Society for Testing and Materials]. Furthermore M5N7 was clearly observed as the strongest of the M lines that disagrees with data presented by Kleykamp [Kleykamp, H. (1981). Wavelengths of the M X-ray spectra of uranium, neptunium, plutonium, and americium. Z Naturforsch36a, 1388–1390], who reported Mβ (M4N6) as the strongest line. Also, after White and Johnson (1970), the line M2N4 should be more intense than M3O5 by a factor of 5. Both our WD and ED spectra show clearly that M3O5 is stronger than M2N4. Altogether, we observed in our WD spectra 26 M lines. In some cases untypical large differences between the line energies given by Bearden [Bearden, J.A. (1967). X-ray wavelengths. Rev Mod Phys39, 78–124] and measured by us were observed.