We developed novel cultures of neonatal rat cardiomyocytes that accurately replicate the MRI-measured microstructure of murine ventricular cross-sections. The individual roles of realistic tissue boundaries and changes in local myofiber direction in impulse conduction were studied by optically mapping membrane voltage (Fig A1 ) in (1) micropatterned cultures with realistic tissue boundaries and natural (anisotropic) cell orientation (Fig A2 ) and (2) controls with the same boundaries but random (isotropic) cell orientation (Fig A3 ). Compared to isotropic controls, anisotropic cultures exhibited a 27% larger spatial dispersion of conduction velocity (CV) at 2 Hz pacing, 17% faster rise in CV dispersion during ramp pacing (5–14 Hz), and 100% higher incidence of remote conduction block, which occurred exclusively at the junctions between the septum and right ventricle (RV) free wall due to current source-sink mismatches and sharp turns in cardiac fiber directions. The incidence of block with ramp pacing was pacing site dependent from the RV, septum, and left ventricle (LV), yielding remote blocks in 77%, 38%, and 0% of anisotropic and 33%, 25%, and 0% of isotropic cultures (Fig B ). In anisotropic cultures, the cycle length of resulting anatomical reentries (normalized by CV) was 34% shorter in LV- vs RV-driven reentry (Fig C1–2 ), due to a shorter wave path. Overall, realistic tissue boundaries and fiber directions synergistically increase the non-uniformity of conduction, render specific regions highly susceptible to conduction block, and determine reentry rate. These purely structural effects are independent of the presence of regional variations of cell ionic properties. This research has received full or partial funding support from the American Heart Association, AHA Mid-Atlantic Affiliate (Maryland, North Carolina, South Carolina, Virginia & Washington, DC).