Nanomedicine and targeted drug delivery systems hold significant promise, but suffer from a lack of successful translation into clinical use. A major factor is the inability to properly characterize important aspects of drug delivery systems in a realistic environment in vitro. Self-assembled particles are of particular concern because aqueous solutions may have different forces contributing to assembly, compared to the complex fluids inside the body. Characterization techniques for nanoparticles, the theory behind the techniques, and their shortcomings for use with blood are surveyed. Previous attempts to replicate or model blood for use with these techniques are examined. The complexity of blood, both its non-Newtonian nature and highly polydisperse cell and protein populations, makes it difficult to reliably study in situ nanoparticle properties in a way that resembles in vivo behavior. However, comparatively little focus is given to this issue: studies frequently use unreliable models of whole blood or jump from aqueous solutions directly to animal models, which can limit translational success. New technologies and methods to study nanoparticles in blood could provide valuable insight into how delivery systems will function in practice, and help to find promising candidates to move into animal or clinical trials.