Hydraulic properties in the near surface play a key role in controlling the complex interactions between atmospheric and terrestrial hydrologic processes. Characteristic curves for water retention and unsaturated hydraulic conductivity modulate the partitioning of precipitation into runoff, evapotranspiration, and groundwater recharge. However, the impact of different methods for estimating these curves on catchment-scale modeling studies is rarely considered. Physically-based hydrologic models are commonly parameterized using pedotransfer functions based on soil texture. These approximations may not be appropriate because soil texture is an indirect proxy for the physical controls on soil-moisture dynamics and the transfer function approach is typically based upon laboratory drainage experiments on repacked core samples. Distributed in-situ measurements of water content, pressure head, and water and energy fluxes reflect the importance of hysteresis and soil structure at larger scales. Inverse methods and model calibration provide alternative approaches for parameter estimation that can make use of distributed measurements to improve simulated representation of hydrologic processes. Results from several field-based investigations of soil-hydraulic properties are presented in conjunction with a sophisticated physics-based model of fully-coupled surface / subsurface flow to demonstrate how different estimation methods can substantially impact simulated hydrologic response at the catchment scale. Results suggest that estimation methods based on a larger support volume, with measurements under both wetting and drying conditions, more accurately incorporate the influence of porous-media structure and heterogeneity on effective hydraulic property values. The implication is that both simulation errors and model uncertainty can be reduced by using commensurate scales of parameter estimation and model application.