Table 1 summarizes the mean numbers of events and interactions of various types which we calculate to occur per single dihydro-orotic dehydrogenase enzyme when irradiated in solution with D37 = 99 Gy of 8.04 keV X-rays under the experimental conditions of Jawad and Watt (1986). There are clearly many orders or magnitude too few direct interactions of X-ray photons, or electrons, with the enzymes for these processes to be responsible for the mean of one inactivating event per enzyme which must occur at the D37 dose. Jawad and Watt (1986) concluded that the enzyme inactivation was predominantly due to direct interaction of an X-ray photon with a non-metal atom of the enzyme, but our analysis shows that this is not possible by five orders of magnitude. Of the possible mechanisms (a)-(d), the only one which remains feasible in this experimental system is the indirect action of radiolysis products from the solution (d). Diffusion distances of the order of 0.1 micron may be quite adequate for such inactivation. The situation would be very different for enzymes or other molecules within mammalian cells where diffusion distances are very much smaller (probably of the order of a few nanometres). Our analysis leaves unexplained the fairly small (20 per cent) change in effectiveness reported by Jawad and Watt (1986) for X-ray energy above, as compared to below, the K-absorption edge of Fe. The experimental observation is directly dependent on the accuracy of the dosimetry at these two energies; we have made no attempt to evaluate this.