The clinical translation of nanomedicine is often hindered by inefficient delivery, with the dense extracellular matrix (ECM) posing a significant physical barrier to transport. A predictive understanding of how nanoparticle (NP) properties and ECM microstructure govern transport is crucial for the rational design of effective nanotherapeutics. This study establishes an integrated computational and experimental framework to systematically investigate these relationships in collagen hydrogels, a primary model system of the ECM. A microstructure-level computational model based on finite element method (FEM) simulations in COMSOL Multiphysics is combined with Brownian dynamics modeling in MATLAB to predict the hydraulic permeability and effective diffusivity in simulated collagen networks with varying collagen concentrations (1.5 - 6.0 mg/mL) and fiber anisotropies (parallel, transverse, random). These predictions were validated against experimental measurements in fabricated hydrogels. Permeability was measured using a suspended perfusion assay, while diffusivity was characterized for solutes ranging from small molecules (Doxorubicin, \({D}_{h}=1.5 {\text{nm}}\) ) to 210 nm NPs using a microfluidic advection-diffusion assay. Both approaches confirmed that increasing collagen concentration significantly impedes transport. While permeability was sensitive to microstructural anisotropy, diffusivity was primarily dictated by solute size and overall matrix density. For large NPs, steric hindrance and interfacial trapping were identified as critical transport limitations. These findings provide a mechanistic basis for understanding transport in dense biological tissues and offer quantitative guidance for designing nanomedicines with enhanced penetration capabilities.