In x-ray diffraction spectrometry as well as in the many other scientific, engineering and medical disciplines involving photon radiation, x-ray attenuation coefficients are required as input data. Within a very few years after the discovery of x-rays by Röntgen in 1895, the transmission of a narrow (parallel) beam of x-rays through layers of different materials was measured and quantified with respect to photon (x-ray) energy and atomic number of the material by Barkla and Sadler in 1907. This quantification is in terms of the mass attenuation (or absorption) coefficient m/r (cm 2 g -1 ) which for monoenergetic photons can be defined as m/r = x -1 ln(I 0 /I ); or I /I 0 = exp[-(m/r)x] in which x is the mass thickness of the layer in units of g cm -2 , I 0 is the intensity (e.g. photons cm -2 s -1 ) of the incident beam, I is the intensity of the transmitted beam (measured with the layer interposed), r is the density of the layer in g cm -3 and m is the linear attenuation (or absorption, m * ) coefficient in cm -1 . Current compilations of m/r are derived from theoretical or semi-empirical values of the most-probable individual processes according to m/r = (s pe + s incoh + s coh + s pair + s trip )/uA in which s pe is the atomic photoeffect cross-section, s incoh and s coh are the incoherent (Compton) and coherent (Rayleigh) scattering cross-sections, respectively, s pair and s trip are the cross-sections for electron-positron pair production (creation) in the field of the nucleus and in the field of the atomic electrons ('triplet' production), respectively, and the factor 1/uA, where u is the atomic mass unit and A is the relative atomic mass of the target element, converts the s i units from cm 2 per atom to cm 2 g -1 . Since s pair and s trip have thresholds above 1 MeV, these do not affect the XRS photon energy region, nor does the photonuclear crosssection s ph.n. which has threshold of 5 MeV or higher and is not included in current m/r compilations. In this brief and arbitrarily selective review, a historical account is given of the evolving compilations of m/r from Barkla and Sadler's work at the beginning of this century, up to the NIST and Livermore compilations as this century comes to its close. For the latter compilations of m/r, some rough estimates of the 'envelope of uncertainty' in different ranges of photon energy are given.