Column experiments were conducted to assess the removal of LNAPLs by aquifer flushing. The LNAPLs included toluene, benzene-toluene mixture, and jet fuel JP-4 in sand and soil columns. Dissolved LNAPL concentrations measured in the column effluent suggested three dissolution periods: the first period, with increasing dissolved effluent concentrations, corresponded to development of the mass transfer zone. The second period, with approximately constant dissolved effluent concentrations, indicated propagation of the mass transfer zone through the LNAPL region. The third period, with decreasing dissolved concentrations, indicated that the length of the LNAPL region was less than that of the mass transfer zone. The third period for all LNAPLs exhibited prolonged tailing and asymptotic concentration decrease to zero. This was attributed to decreasing interfacial areas and mass transfer limitations from within the LNAPL mass. Dissolved concentrations produced from a multicomponent LNAPL were lower than those from pure LNAPLs, as expected from thermodynamics. After 132 pore volumes, toluene removal in the single LNAPL experiment was 98.6%. After 53 pore volumes, benzene and toluene removal in the two LNAPL system was 94.6% and 88.3%, respectively. A simple procedure was used to estimate the volume of groundwater required to flush individual jet fuel components from the aquifer material, and the results were compared with the column experimental data.