Together, resolution and dynamics represent the two linked frontiers in the field of single particle reconstruction. Subnanometer resolutions have now been demonstrated numerous times, and several specific structures have now moved past the 5 A resolution mark. Yet these resolutions remain elusive for a wide range of systems with high biological impact, often due to the inherent flexibility of the assemblies in question. However, this fact can also work to our advantage, as, when present, structural flexibility almost certainly plays an important role in the biological function of the assembly. Since single particle analysis operates on images of individual proteins/assemblies, it is possible to computationally subclassify particles based on their conformational state. This process also has the side-effect of making each classified particle set more structurally homogeneous, and thus potentially improving the resolution of the individual reconstructions. The largest limitation of this approach is that the motions performed by many systems can be extremely complex, and thus difficult to fully characterize. This talk will present some new ideas for approaching the problem of structural heterogeneity and investigate the analyses that are possible in reconstructions at better than 5 A resolution.