Dye-sensitized solar cells (DSCs, also known as Gratzel cells) mimic the photosynthetic process by using a sensitizer dye to harvest light energy to generate electrical power. Several functional features of these photochemical devices are unusual, and DSC research offers a rewarding arena in which to test new ideas, new materials, and new methodologies. Indeed, one of the most attractive chemical features of the DSC is that the basic concept can be used to construct a range of devices, replacing individual components with alternative materials. Despite two decades of increasing research activity, however, many aspects of the behavior of electrons in the DSC remain puzzling. In this Account, we highlight current understanding of the processes involved in the functioning of the DSC, with particular emphasis on what happens to the electrons in the mesoporous film following the injection step. The collection of photoinjected electrons appears to involve a random walk process in which electrons move through the network of interconnected titanium dioxide nanoparticles while undergoing frequent trapping and detrapping. During their passage to the cell contact, electrons may be lost by transfer to tri-iodide species in the redox electrolyte that permeates the mesoporous film. Competition between electron collection and back electron transfer determines the performance of a DSC: ideally, all injected electrons should be collected without loss. This Account then goes on to survey recent experimental and theoretical progress in the field, placing particular emphasis on issues that need to be resolved before we can gain a clear picture of how the DSC works. Several important questions about the behavior of "sticky" electrons, those that undergo multiple trapping and detrapping, in the DSC remain unanswered. The most fundamental of these concerns is the nature of the electron traps that appear to dominate the time-dependent photocurrent and photovoltage response of DSCs. The origin of the nonideality factor in the relationship between the intensity and the DSC photovoltage is also unclear, as is the discrepancy in electron diffusion length values determined by steady-state and non-steady-state methods. With these unanswered questions, DSC research is likely to remain an active and fruitful area for some years to come.