The LambdaCDM model has enjoyed considerable success in describing the clustering of mass, on large scales. On smaller scales, in regions of high DM density, there have been tantalizing suggestions of departures from the predictions of the simplest cold dark matter models. In particular, CDM predictions for halo substructure, central halo profiles and concentrations, and halo triaxiality have all presented possible discrepancies with observations. In this talk, I will discuss how gravitational lensing, and in particular strong lensing, may be used as a probe of these three aspects of dark matter halos for objects ranging in mass from 10^8 Msun up to the most massive galaxy clusters. I will give the most recent constraints derived from lensing analyses, and lastly will discuss future prospects for the use of lensing to study dark matter.