Laboratory column tests have been carried out to assess the transport behaviour of Mecoprop in Mercia Mudstone clay and Oxford Clay. Artificially consolidated clay samples were compressed in a triaxial cell to stresses representative of those at the base of a landfill. Uniform steady flow was achieved, and there was no evidence of ‘bypass’ flow in one or more fast streamtubes. Analysis of Mecoprop and bromide breakthrough curves showed that the transport characteristics were linear (within noise), with no evidence of irreversible sorption. The possibility of dual-porosity or kinetic sorption processes was not conclusively eliminated. Truncated temporal moments allowed estimates of the linear retardation factor for Mecoprop, which were shown to be a lower bound. Based on modelling, the retardation factor of Mecoprop in Oxford Clay was estimated to be 17.1, compared with 65.4 in batch sorption tests. The mean retardation factor for Mecoprop in Mercia Mudstone was 3.6, whereas the value from batch sorption tests was 10.0. These data confirm that retardation factors calculated from sorption isotherms obtained from batch experiments on disaggregated samples substantially overestimate the retardation likely to be observed in compacted clay liners; therefore values from batch tests should be used with caution in groundwater risk assessments.