Compressive sensing (CS) is a framework in which one attempts to measure a signal in a compressive mode, implying that fewer total measurements are required vis à vis direct sampling methods. Compressive sensing exploits the fact that the signal of interest is compressible in some basis, and the CS measurements correspond to projections (typically random projections) performed on the basis function coefficients. In this paper, we demonstrate that ideas from compressive sensing may be exploited in the context of electromagnetic modeling, here multi-static scattering from an arbitrary target. In this context, the computational analysis may be viewed as a numerical experiment, and ideas from compressive sensing may be used to reduce the number of computations required for target characterization. It is demonstrated that the compressive sensing framework may be applied with relatively minor modifications to many existing numerical models, with examples presented here for a fast-multipole computational engine. © 2009 Elsevier Inc. All rights reserved.