The mineral content of dental enamel is commonly measured by X-ray attenuation experiments. Most studies have used contact microradiography in which intensities are measured with photographic film which is convenient and gives high spatial resolution. However photon counting intensity measurements are to be preferred in many experiments (longitudinal and scanning microradiography, and microtomography), as illustrated here, because they have a larger dynamic range and greater sensitivity to small intensity changes. Additionally, the detector and specimen are well separated which allows the pseudo-continuous study of de- and remineralization. The mineral content is often quoted as 95 wt% or 87 vol% hydroxyapatite for permanent human enamel. This determination from attenuation experiments requires accurate values of elemental mass attenuation coefficients and a number of assumptions. The effects of possible choices of these are considered and it is shown that the most important is the density of enamel mineral used in conversion of wt% to vol%. If the density is taken as 2.99 g cm(-3), as recently suggested (J.C. Elliott, Dental Enamel, Ciba Foundation Symposium 205, Wiley, Chichester, pp. 54-72, 1997), instead of 3.15 g cm(-3) as for hydroxyapatite, the calculated vol% is approximately 93 instead of approximately 87.