The design of in-situ remediation is currently based on a description at the macroscopic scale. Phenomena at the pore and pore-network scale are typically lumped in terms of averaged quantities, using empirical or ad hoc expressions. These models cannot address fundamental remediation issues at the pore and pore network scales, including: The emplacement in-situ of the contaminant NAPL, and the displacement patterns; the mass transfer of the contaminant from the NAPL to the groundwater or from the groundwater to a sparging fluid, and of the remedial agents to the NAPL; and possible micro-scale flow instabilities during injection of a remedial fluid. The objective of this work is to obtain a fundamental understanding by conducting theoretical, experimental and computational pore-scale studies. Emphasis is placed at the pore network scale. Use of this information can be incorporated in macroscopic simulators to provide fundamentally correct expressions for the various coefficients or parameters, currently treated empirically. The theoretical findings are compared with findings from experiments in glass micro-models and Hele-Shaw cells.