Context: Binaries and multiple systems are very frequent and form large fraction of all stellar systems. In contrast to their single counterparts, studying binaries provides the possibility to accurately determine fundamental properties of their components that are needed for testing models of stellar structure and evolution. On top of that, binaries can be used for accurate distance determinations. The mass exchange in close binaries remains the only mechanism, which completely alters their evolution. Aims: The primary goal of my doctoral study was to determine orbital elements of selected systems and properties of their components - masses, radii, and ef- fective temperatures. - In case of more complicated objects (e.g. interacting multiple systems, mass-transferring binaries, . . . ) the secondary goal was to confront our results with predictions of theoretical models. Methods: Studies that I co-authored were based on three different types of ob- servations, each sensitive to partly different properties of studied systems - photometry, spectroscopy, and spectro-interferometry. The analysis was carried out through several "observation-specific" models, whose outcome was critically compared to each other and to previous studies of the object in question. Results: Throughout my study I contributed to: (i)...