Hygroexpansion, CP/MAS 13C-NMR, WAXS and SAXS measurements were carried out on sheets made from four different commercial pulps of varying lignin content. Non-directional laboratory sheets were made at different press levels from the pulps following different degrees of beating. The sheets were dried both freely and with restraints. Measurements were made on sheets before and after moisture cycling to determine hygroexpansion coefficients, changes in cellulose average lateral fibril dimensions and average cellulose crystallite sizes, with the aim of understanding macroscale and nanoscale changes as the result of moisture cycling. Within the sheets consistent and statistically significant structural changes were observed on both macro and nanoscale. On the macroscale, moisture cycling consistently induced irreversible shrinkage in sheets dried with restraints, but less so in the case of sheets dried freely. The hygroexpansion coefficients were typically higher for freely dried sheets compared with sheets dried with constraints. On the nanoscale, moisture cycling consistently caused an increase in the average crystallite sizes (WAXS) and the average lateral fibril dimensions (CP/MAS 13C-NMR), though the latter occurred with poor statistical significance. These changes were interpreted as an increase in the degree of order in the cellulose fibril interior/cellulose crystallite. There were no profound differences in the nanoscale changes observed for sheets dried with restraints and for sheets dried freely. Changes in the fibre wall nanostructure were of similar magnitudes when comparing results from freely dried low grammage sheets (less abundant inter-fibre joints) with freely dried sheets of higher grammage (more abundant inter-fibre joints). No obvious correlations were found between the macroscale and nanoscale measurements. The proposed explanation for this was that the nanoscale structural changes occurred similarly throughout the entirety of the fibre wall, independent of the proximity to an inter-fibre joint, and that the nanoscale structural changes were mainly the result of water penetrating into the interior of cellulose fibril aggregates. By using the same fibril model for evaluation of CP/MAS 13C-NMR and WAXS data, good-to-reasonable agreement were found for estimates of the degree of cellulose crystallinity.