Cardiac architecture is inherently three-dimensional, yet most characterizations rely on two-dimensional histological slices or dissociated cells, which remove the native geometry of the heart. We previously developed a method for labeling intact heart sections without dissociation and imaging large volumes while preserving their three-dimensional structure. We further refine this method to permit quantitative analysis of imaged sections. After data acquisition, these sections are assembled using image-processing tools, and qualitative and quantitative information is extracted. By examining the reconstructed cardiac blocks, one can observe end-to-end adjacent cardiac myocytes (cardiac strands) changing cross-sectional geometries, merging and separating from other strands. Quantitatively, representative cross-sectional areas typically used for determining hypertrophy omit the three-dimensional component; we show that taking orientation into account can significantly alter the analysis. Using fast-Fourier transform analysis, we analyze the gross organization of cardiac strands in three dimensions. By characterizing cardiac structure in three dimensions, we are able to determine that the α crystallin mutation leads to hypertrophy with cross-sectional area increases, but not necessarily via changes in fiber orientation distribution.