The possible effects of paramagnetic relaxation on the apparent volumes of activation for exchange reactions in solution, as measured by NMR at high pressures, are considered Two model paramagnetic systems that do not undergo ligand exchange on the NMR time scale were examined: tri(acetylacetonato)chromium(III) in various perdeuterated solvents, and tris(ethylenediamine)nickel(II) ion in ethylenediamine solvent. No pressure dependence was discernible up to 200 MPa for the chemical shifts of H-1 (exemplifying nuclei of spin 1/2) in the Cr(III) complex, or of solvent N-14 (representing quadrupolar nuclei) in the Ni(II)-ethylenediamine case. The line widths Delta nu(1/2), however, were significantly dependent on pressure. For H-1 in the Cr complex, the increase of Delta nu(1/2) with pressure was less than expected from the theory of scalar interactions, and was small enough to imply that any contribution from this source to the observed volume of activation in exchanging systems may be neglected. For N-14 in liquid ethylenediamine, the increase of Delta nu(1/2) with pressure was significantly greater when a paramagnetic solute was present. Thus, before the observed Delta nu(1/2) at a pressure P (in MPa) measured by the NMR of a quadrupolar nucleus can be used to obtain a chemical exchange rate for a paramagnetic solute, it should be reduced by an amount Delta nu(1/2)(0) exp{theta(P-0.1)(Delta V-V(double dagger)/RT+kappa)}, where Delta V-V(double dagger) is the activation volume for viscous flow and kappa the compressibility of the solvent, Delta nu(1/2)(0) is the linewidth at 0.1 MPa in the absence of chemical exchange, and theta is a scaling factor between 0 and 1. The factors Delta nu(1/2)(0) and theta(Delta V-V(double dagger)/RT+kappa) are obtainable from measurements with a chemically equivalent, nonexchanging, paramagnetic solute.