Purpose: This work introduces a new lattice geometry library, egs_lattice, into the EGSnrc Monte Carlo code, which can be used for both modeling very large (previously unfeasible) quantities of geometries and establishing recursive boundary conditions. The reliability of egs_lattice, as well as EGSnrc in general, is cross-validated and tested... Methods: New Bravais, cubic, and hexagonal lattice geometries are defined in egs_lattice and their transport algorithms are described. Simulations of cells and Gold NanoParticle (GNP) containing cavities are implemented to compare to published Geant4-DNA and PENELOPE results. Recursive boundary conditions, implemented through a cubic lattice, are used to perform electron Fano cavity tests. Results: Lattices are successfully implemented in EGSnrc. EGSnrc calculated doses to cell cytoplasm and nucleus when irradiated by an internal electron source with a median difference of 0.6% compared to Geant4-DNA. EGSnrc calculated the ratio of dose to a microscopic cavity containing GNPs over dose to a cavity containing a homogeneous mixture of gold, and results generally agree (within 1%) with PENELOPE. The Fano test is passed (sub-0.1%) for all energies/ cells considered. Additionally, the recursive boundary conditions used for the Fano test provided a factor of over a million increase in efficiency in some cases. Conclusions: The egs_lattice geometry library, currently available as a pull request on the EGSnrc GitHub develop branch, is now freely accessible as open-source code. Lattice geometry implementations cross-validated with independent simulations in other MC codes and verified with the electron Fano cavity test demonstrate not only the reliability of egs_lattice, but, by extension, EGSnrc's ability to simulate transport in nanometer geometries and score in microscopic cavities. (read more)