Inclusion of solvent effects in biomolecular simulations is most ideally done using explicit methods, as they are able to capture the heterogeneous environment typical of biomolecules and systems involving them (e.g., proteins at solid interfaces). Common explicit methods based on molecular solvent models (e.g., TIP and SPC models) and molecular dynamic or Monte Carlo simulation are computationally expensive and are, therefore, not well-suited to situations where many simulations are required (e.g., in the ab initio structure prediction or design contexts). In such cases, more coarse-grained explicit approaches such as the Langevin dipole (LD) method of Warshel and co-workers are more appropriate. The recent incarnations of the LD method appear to produce good solvation free energy estimates. These incarnations use charges and solute structures obtained from high-level quantum mechanics simulations. As such an approach is clearly not possible for larger solutes or when many structures are to be considered, an alternative must be sought. One possibility is to use structures and charges derived from an existing analytical potential modelwe report on such a coupling here with the Amber potential model. The accuracy and computational performance of this hybrid approach, which we term LD-Amber to distinguish it from previous incarnations of the LD method, was assessed by comparing results obtained from the approach with those from experiment and other theoretical methods for the solvation of 18 amino acid analogues and the alanine dipeptide. This comparison shows that the LD-Amber approach can yield results in line with experiment both qualitatively and quantitatively and is as accurate as other explicit methods while being computationally much cheaper.