Mass spectrometry has in recent years been established as the standard method for protein identification and characterization in proteomics with excellent intrinsic sensitivity and specificity. Fourier transform ion cyclotron resonance is the mass spectrometric technique that provides the highest resolving power and mass accuracy, increasing the amount of information that can be obtained from complex samples. This thesis concerns how useful information on proteins of interest can be extracted from mass spectrometric data on different levels of protein structure and how to obtain this data experimentally. It was shown that it is possible to analyze complex mixtures of protein tryptic digests by direct infusion electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry and identify abundant proteins by peptide mass fingerprinting. Coupling on-line methods such as liquid chromatography and capillary electrophoresis increased the number of proteins that could be identified in human body fluids. Protein identification was also improved by novel statistical methods utilizing prediction of chromatographic behavior and the non-randomness of enzymatic digestion. To identify proteins by short sequence tags, electron capture dissociation was implemented, improved and finally coupled on-line to liquid chromatography for the first time. The combined techniques can be used to sequence large proteins de novo or to localize and characterize any labile post-translational modification. New computer algorithms for the automated analysis of isotope exchange mass spectra were developed to facilitate the study of protein structural dynamics. The non-covalent interaction between HIV-inhibitory peptides and the oligomerization of amyloid β-peptides were investigated, reporting several new findings with possible relevance for development of anti-HIV drug therapies and understanding of fundamental mechanisms in Alzheimer’s disease.