Dye-sensitized solar cells (DSCs) constitute a novel class of hybrid organic-inorganic solar cells. At the heart of the device is a mesoporous film of titanium dioxide (TiO2) nanoparticles, which are coated with a monolayer of dye sensitive to the visible region of the solar spectrum. The role of the dye is similar to the role of chlorophyll in plants; it harvests solar light and transfers the energy via electron transfer to a suitable material (here TiO2) to produce electricity — as opposed to chemical energy in plants. DSCs are fabricated of abundant and cheap materials using inexpensive processes (e.g. screen-printing) and are likely to be a significant contributor to the future commercial photovoltaic technology portfolio. The work conducted during this thesis aimed at optimizing the DSC using three different strategies: The use of versatile organic sensitizers for stable and efficient DSCs, the study of tandem device architectures in combination with other solar cells to harvest a larger fraction of the solar spectrum, and the development of a validated optoelectric model of the DSC. Organic donor-π-acceptor dyes are an interesting alternative to the standard metal-organic complexes used in DSCs. Efficient photovoltaic conversion and stable performance could be demonstrated with three classes of donor systems, namely diphenylamine, difluorenylaminophenyl, and π-extended tetrathiafulvalene. The highest conversion efficiencies were obtained with a difluorenylaminophenyl donor system (η = 8.3 % with a volatile electrolyte and η = 7.6 % with a solvent-free ionic liquid, which was a new record for organic dyes at the time of publication). Surprisingly, efficient regeneration of the oxidized dye by the I-/I3- redox mediator was found with the π-extended tetrathiafulvalene system, even though the thermodynamic driving force was as low as 150 mV. So far driving forces of 300-500 mV had been regarded as necessary for efficient regeneration of the dye cation. Also, important structure-property relationships pertaining to the recombination of electrons with the electrolyte and to the stability of the device could be identified (i.e. effect of linear vs. branched structure, linker length, and moieties used). The power conversion efficiency of solar cells can be extended beyond the limit for a single cell (∼ 30 %) by using multiple cells with different optical gaps in a tandem device. DSCs and chalcopyrite Cu(In,Ga)Se2 (CIGS) solar cells have complementary optical gaps and are thus suitable systems for integration in a tandem device. It was shown that a monolithic DSC/CIGS tandem device has the potential for increased efficiency over a mechanically stacked device due to increased light transmission to the bottom cell, and a monolithic DSC/CIGS device with an initial efficiency of η = 12.2 % was demonstrated. The degradation of the devices — induced by the corrosion of the CIGS cell in contact with the I-/I3- redox mediator — could be retarded with a protective thin conformal ZnO/TiO2 oxide layer coated on the CIGS cell by atomic layer deposition. Finally, an experimentally validated optical and electrical model of the DSC has been developed to assist the optimization process, which is predominantly conducted by empirical means in the DSC research community. The optical model allows to accurately calculate the internal quantum efficiency of devices, i.e. the ratio of extracted electrons to absorbed photons by the dye, a crucial and so far difficult to determine characteristic. Intrinsic parameters — like injection efficiency, electron diffusion length, or distribution of trap states in the TiO2 — can be extracted from experimental steady-state and time-dependent data with the electric model. The model allows to make a comprehensive and quantitative loss analysis of the different optical and electric loss channels in the DSC. The model has been implemented with a graphical user interface for straightforward usage. All three optimization strategies — organic dyes, tandem architecture, and device modeling — developed during this thesis make a valuable contribution to the development and commercialization of inexpensive and high efficiency DSCs. They enable a comprehensive view of the system and pave the way for a systematic analysis and reduction of losses, which has been the ultimate route to success for several established photovoltaic technologies.