Estimating material properties of natural materials is a goal shared between near-surface exploration, infrastructure management, and even conservation of historic sites. Existing methods (e.g., seismic, ultrasonics, and direct testing) for characterizing these properties of interest are not yet available at the desired scale or configuration. GPR is an established method for characterizing the physical configuration of these sites by mapping subsurface features. A new analysis method applying seismic and image processing attribute analysis to GPR data has promising results for material identification and possibly characterization. To further explore this capacity, an experiment to characterize concrete laboratory specimens using GPR was performed. Creating a unique data set, GPR scans and attribute analyses are paired with direct testing of the samples. A variety of attributes and regression analyses have been able to predict porosity of the samples from the mean instantaneous GPR amplitudes with some success (R2 = 0.8). Density and compressive strength are more difficult to predict from the GPR data, but porosity is an excellent proxy property that can be related to other physical properties using secondary relationships. By continuing to explore these relationships and test them on GPR data for other materials (stone, brick, etc.), the results could be extended and refined to provide fine-scale estimation of important properties to characterize reservoirs, basins, and other geological or anthropogenic features of interest.