The development of strategies for optimizing dye-sensitized solar cells (DSC) has been hampered by lack of understanding of some of the fundamental processes involved in cell operation. Whereas the injection and collection of electrons at short circuit is evidently not a problem, the factors that determine the achievable open circuit voltages and fill factors are less clear. The description of electron transfer from the TiO₂ to the redox electrolyte of hole conducting medium is not well understood, and the possible role of surface states remains to be established. Attempts to optimize cells have often been based on a misinterpretation of the transient photocurrent and photovoltage response of DSCs. For example, changes in electrolyte composition or in the adsorbed dye are reported as ‘increasing the electron lifetime’ or ‘enhancing electron transport’. In fact, the effects may arise simply from changes in the conduction band energy of the TiO₂. The transient photocurrent and photovoltage response are strongly influenced by electron trapping, so that the actual values of the diffusion coefficient and lifetime of free electrons can only be found by assuming a model. Unfortunately, current models are clearly too simple to describe the behaviour of real cells. Questions arise concerning the role of surface sates in electron transfer and of possible voltage dependent barriers to electron extraction at the substrate in determining fill factors. This lecture explores some of these issues and shows how a self-consistent experimental approach combined with new modelling results can help give a clearer understanding of DSC function.