The membrane of the endoplasmic reticulum (ER) is the site of numerous complex activities essential to the survival of eukaryotic cells. In the membrane sheets of the rough ER (rER), integral membrane proteins are synthesized, folded, modified, and transported throughout the cell. In the smooth ER (sER) tubules, lipids are synthesized and processed with a variety of functionalities, and trafficked to specific membrane compartments. The sER is also the site of calcium storage and controlled release, an important function for several cell types including neurons and muscle. The sER requires energy and the action of numerous proteins to maintain its highly curved, tubular shape and its reticular, interconnected structure. The recent discovery of a family of proteins called atlastins has helped answer some of the questions surrounding the tubular, reticular nature of the sER. Atlastins comprise a group of ER resident proteins that have been shown to facilitate the fusion of membrane tubules within this organelle. They are part of the dynamin superfamily of proteins, which use the energy stored in GTP to sculpt membranes throughout the cell. However, the exact molecular mechanism for how atlastin mediates membrane fusion remained mysterious. In this study, we have taken apart and analyzed atlastin-1, one of three isoforms of atlastin in humans. This isoform is found primarily in the central nervous system and is mutated in the neurodegenerative disorders Hereditary Spastic Paraplegia and Hereditary Sensory Neuropathy. This study has resulted in a collection of three-dimensional, high-resolution structures of atlastin-1's catalytic core fragment, comprising its GTPase and middle domains bound to various forms of the guanine nucleotide. These structures revealed key information about the catalytic mechanism of atlastin-1, as well as interesting and important conformational changes that occur within the structure. Using these structures, as well as information from small-angle X-ray scattering (SAXS), sizeexclusion chromatography coupled to multi-angle light scattering (SEC-MALS), Förster resonance energy transfer (FRET), and enzymatic activity assays, we have put together a general working model for atlastin membrane fusion and have gathered important information about its mechanism that may be used to target the protein for the treatment of disease.