Polymer hydrogels play a key role as sieving matrix materials for electrophoresis of charged biomolecules like DNA and proteins. In order to rationally select hydrogels that possess properties optimally suited for these applications, it is important to understand the nature of the gel pore structure and the parameters through which it can be manipulated. The characteristic average pore size and degree of polydispersity are particularly important quantities but they have proven difficult to measure experimentally because most hydrogels are mechanically fragile and easily damaged. We have recently developed an approach that uniquely combines complementary information from in-situ dynamic mechanical analysis, thermoporometry, and transmission electron microscopy in a way that enables both the mean pore size and the pore size distribution to be quantitatively characterized in polymer hydrogels. Here, we apply these techniques to investigate how chemical composition (relative amounts of monomer and crosslinker) and polymerization conditions (chemical versus UV initiation chemistry) influence the nanoscale pore morphology in polyacrylamide/bis-acrylaimde hydrogels. When compared with corresponding measurements in agarose gels, these measurements provide insights into the interplay between the nanoporous architecture of the gel matrix and electrophoretic separation behavior. This information also allows us to identify synthesis conditions that yield hydrogels with comparable mean pore sizes, but whose pore size distributions incorporate different degrees of polydispersity. The critical role played by polydispersity becomes evident when we examine controlled release of macromolecules from these gels, where we observe dramatically altered release kinetics in gels with different polydispersity that would otherwise appear identical if compared on the basis of their mean pore size alone.