The study of three-dimensional protein structure is essential to understanding the myriad of functions of these important biomolecules and the complexes they form. Viruses are entities that exploit their protein machinery with a high degree of efficacy and economy and the structural characterisation of virus proteins is vital to discovering new ways to combat disease. No single biophysical technique can provide a complete picture of protein architecture, so a range of complementary approaches is required. This thesis presents the use of noncovalent mass spectrometry (MS) techniques for the study of a diverse range of viral structural proteins, including human, animal, and bacterial viruses. Noncovalent MS and travelling-wave ion mobility spectrometry-MS (TWIMS-MS) were used successfully to analyse the dynamic behaviour of the assembly domain (CP149) of the capsid protein of hepatitis B virus (HBV), and, along with limited proteolysis, locate these dynamics to the C-terminus of a half-dimer. The C-terminus is crucial for virus particle assembly and such direct conformational analysis is • difficult with more established techniques. Following this, investigations were made into the effects of an anti-HBV small molecule, HAP-l, on the behaviour of CP149. The molecule had no observable effect on CP149 conformation but interestingly, induced the formation of oligomers which were found by TWIMS-MS to have a globular structure, i.e. not 'sheet-like' as may be expected for assembly intermediates. The small molecule also promoted the formation of large flat particles observed by electron microscopy (EM). The formation of these two rather different classes of particle possibly reflects two facets of the activity of the HAP class of compounds in cells. The technique of collision-induced unfolding (ClU) was studied to investigate whether this can be used to characterise and differentiate between CP149 mutants. Although valuable insights into the unfolding pathway were revealed, together with a number of distinct conformational families, it was found that ClU could not distinguish between the mutations. This is thought to be a result of the substitution of only hydrophobic residues, which may not influence gas-phase structure due to the absence of the hydrophobic effect. MS, TWIMS-MS, EM and coarse-grained structural modelling were utilised to study the oligomerisation processes of the nucleocapsid (N) proteins of Bunyamwera and Schmallenberg viruses from the Bunyaviridae family, a topical and increasingly important class of viruses, for which there is currently no high resolution structural information. The N proteins were shown to oligomerise on binding to RNA to form a range of well-defined but different oligomers. The MS and EM data suggest the shapes of the oligomers to be based on ring structures rather than possible tetramer-based arrangements. These data represent the first insights into the assembly pathways of these important viruses, and the first such MS assembly study of a non-spherical virus class. Noncovalent MS and TWIMS-MS were used to study the baseplate complexes of bacteriophages known to infect the bacterium lactococcus lactis, which is important to the dairy industry. The phages p2 and TP901-1 have multiprotein organelles termed baseplates that mediate cell attachment and genome release and are critical for infection. The complexes, of 0.8 and 1.8 MDa, respectively, were characterised in terms of their mass, and stoichiometry as were several subcomplexes likely to be important in assembly. The studies, in conjunction with published EM and X-ray crystallography data, highlighted the importance of the distal tail protein as a central hub to control baseplate assembly and showed that valuable structural information can be gained in the absence of high resolution crystallography data. In summary, MS and TWIMS-MS were found to be valuable tools with which to study protein dynamics, structure and assembly, which makes them vital for the advancement of biophysical characterisation of viral systems.