Part of a study to assess the feasibility of coal tar site remediation by the injection and recovery of water‐miscible, biodegradable solvents is described. The overall objective of the study was to obtain estimates of cleanup times and solvent volumes required for a hypothetical contaminated site. We describe the results of column experiments, performed by passing solvent‐water solutions through glass bead‐packed columns contaminated with coal tar blobs at low volumetric saturation. This experimental design allowed the investigation of dissolution rate limitations at residual saturations similar to those observed in the field, albeit under more homogeneous conditions. The column effluent data were modeled with a modified advection‐dispersion equation that includes a mass‐transfer source term for coal tar dissolution. This source term empirically relates mass transfer to experimental variables such as coal tar saturation and solvent flow velocity using fitting parameters. Values of parameters in the mass‐transfer source term were estimated from experimental data. The modeling approach and the parameter values developed in this study have been used in simulations of thein situsolvent extraction process.