Understanding and controlling water in wood is critical to both improving forest products moisture durability and developing new sustainable forest products-based technologies. While wood is known to be hygroscopic, there is still a lack of understanding of the nanoscale wood-water interactions necessary for increased moisture-durability and dimensional stability. My PhD thesis focuses on the development and implementation of neutron scattering methods that can provide insight on both the structural and dynamical changes associated with these interactions so that products with improved moisture durability can be developed efficiently. Using small angle neutron scattering (SANS) and a custom-built in situ relative humidity chamber I studied the anisotropic moisture-induced swelling of wood nanostructure. First, I studied the effects of sample preparation by comparing SANS patterns of wiley milled wood and intact latewood cell walls, and found that scattering from intact wood provides more information about the spatial arrangement of the wood nanostructures inside the cell wall. Comparisons between SANS patterns from earlywood and latewood, show that the higher cell wall density of latewood results in patterns with more pronounced anisotropic features. Then, by measuring latewood loblolly pine sections obtained from the same growth ring and prepared in each of the primary wood planes, I tracked the cellulose elementary fibril spacing as a function of humidity in both intact and partially cut cell walls. These studies showed that even though swelling at the elementary fibril spacing is responsible for the majority of the transverse swelling observed at the S2 level, it is not primary plane dependent. Additionally, there were no differences in the elementary fibril spacing between partially-cut and intact cell walls, except at high humidity where the spacing in partially-cut cells was higher. SANS was also used to study the effects of two chemical modifications, namely, adhesive infiltration and and acetylation, on the wood nanostructure as well as its moisture-induced swelling. Tangential-longitudinal latewood loblolly pine 0.5 mm thick sections were acetylated or treated with an adhesive (Phenol-formaldehyde (PF) or polymeric methylene diisocyanate (pMDI)) using deuterated or hydrogenated chemicals. Contrast variation experiments on wood modified with deuterated chemicals revealed that PF can infiltrate the regions between the elementary fibrils, while acetylation does not. The moisture-induced swelling of the chemically modified wood was investigated by studying the samples modified with hydrogenated chemicals using SANS and the previously built humidity chamber. These studies revealed that while both PF and pMDI can infiltrate the microfibrils, only PF reduced significantly the swelling at both the elementary fibril and bulk levels. In acetylated samples, the elementary fibril spacing was proportional to the moisture-content of the sample, which was reduced with increasing acetylation. This suggested that the acetylation treatment did not reduce the swelling at the elementary fibril but prevented water from entering the microfibril by modifying the regions surrounding the elementary fibrils. Using quasi-elastic neutron scattering (QENS) and a custom-built in situ relative humidity sample environment I measured experimentally the (5 – 400 ps) water dynamics inside wood cell walls for the first time and found that there are two types of bound water in the cell wall, namely, slow and fast water. The motion of both water types is well described by a jump-diffusion model, which corresponds to water molecules whose movement follows a stop and go process. Here, the slow water corresponds to water molecules that are highly associated to the wood polymers, whereas the fast water corresponds to water confined inside nanopores within the wood cell wall.