Multiply scattered waves are often neglected in imaging methods, largely because of the inability of standard algorithms to deal with the associated non-linear models. This paper shows that by incorporating a known environment into the background model, we not only retain the benefits of imaging techniques based on linear models, but also obtain different views of the target scatterer. The net result is an enhanced angular resolution of the target to be imaged. We carry out our analysis in the context of high-frequency radar imaging, in which a steerable beam from a moving platform is used to produce an image of a region on the earth's surface (the target scatterers being buildings, etc). We consider the case where the target we want to image is situated in the vicinity of an a priori known reflecting wall. This is one of the simplest possible multipathing environments for the scatterer, and in the case when the illuminating beam is narrow enough to isolate different scattering paths, we will show that the imaging process achieves enhanced angular resolution. Although we carry out our analysis here in the context of radar, our technique is general enough that it can be adapted to many imaging modalities, such as acoustics, ultrasound, elasticity, etc. The extension of the method to other more complicated environments is also possible. © 2006 IOP Publishing Ltd.