X-ray diffraction experiments on amorphous Fe₄₀Ni₄₀P₂₀ were performed in association with a detailed study of the accuracy, limitations and reproducibility of experimental radial distribution functions (RDF's). A novel counting strategy is presented, and the merits of conducting two diffraction experiments in different geometries as a test for internal consistency are argued. Experimental methods and data-processing techniques are treated in detail. The results indicate that the RDF can be determined virtually with pre-selected accuracy. The positions of the co-ordination shells for Fe₄₀Ni₄₀P₂₀ show a remarkable periodicity.

Comments made by Hajdu & Herms [J. Appl. Cryst. (1986). 19, 198-199] on the original paper by Thijsse [J. Appl. Cryst. (1984). 17, 61-76] are replied to.