Abstract: Using 31P solid-state NMR spectroscopy, anisotropy in the indirect 199Hg-31P spin–spin coupling tensor (ΔJ) for powdered [HgPCy3(OAc)2]2 (1) has been measured as 4700±300Hz. Zeroth-order regular approximation (ZORA) density functional theory (DFT) calculations, including scalar and spin-orbit relativistic effects, performed on 1 and a series of other related compounds show that ΔJ(199Hg, 31P) arises entirely from the ZORA Fermi-contact–spin-dipolar cross term. The calculations validate assumptions made in the spectral analysis of 1 and in previous determinations of ΔJ in powder samples, namely that J is axially symmetric and shares its principal axis system with the direct dipolar coupling tensor (D). Agreement between experiment and theory for various 199Hg, 31P spin–spin coupling anisotropies is reasonable; however, experimental values of 1 J(199Hg, 31P)iso are significantly underestimated by the calculations. The most important improvements in the agreement were obtained as a result of including more of the crystal lattice in the model used for the calculations, e.g., a change of 43% was noted for 1 J(199Hg, 31P)iso in [HgPPh3(NO3)2]2 depending on whether the two or three nearest nitrate ions are included in the model. Finally, we have written a computer program to simulate the effects of non-axial symmetry in J and of non-coincidence of the J and D on powder NMR spectra. Simulations clearly show that both of these effects have a pronounced impact on the 31P NMR spectrum of 199Hg–31P spin pairs, suggesting that the effects should be observable experimentally if a suitable compound can be identified. [Copyright &y& Elsevier]