Photocatalytic (PC) solar hydrogen production is a promising way to provide green hydrogen using only sunlight and abundant reactants such as water. PC approaches use catalytically active semiconductor particles suspended in liquid electrolytes. The particles absorb photons with sufficient energy from sunlight to produce excited charge carriers, which are separated and transported to catalytically active sites where they are used in reduction and oxidation reactions. PC water splitting can be classified in one-step or Z-scheme approaches, where for the latter, typically membranes are employed to separate the product gases (hydrogen and oxygen). This thesis provides a holistic study of PC water splitting by addressing material selection and reactor design challenges through modeling, experiments and techno-economic analysis. PC water splitting research has been focused on the synthesis of photocatalyst materials and their characterization. The material-related challenges are being tackled abundantly whereas little attention has been paid to the conceptual design of reactors, the effect of their design and coupled transport on the material choice and requirements. It is unclear how the conceptual design of reactors affects the overall performance, the limiting efficiencies, and the best choice of materials. To improve our theoretical understanding of the effect of reactor design choice, a transient zero-dimensional two-particle model was developed, accounting for ideal solar absorption, ideal charge generation and transport, the kinetic characteristics at the catalytic sites of the photocatalysts and species transport and mass conservation in the electrolyte. Different conceptual reactor designs (single and dual particle suspensions) were assessed and the sensitivity of the reactor performance on component choice, reactor design, and operating conditions are quantified. The developed model provided theoretical guidelines and limiting efficiencies for photocatalyst particle-suspension reactors, offered a feasibility assessment and allowed to quantify the performance potential of various reactor designs. To increase the accuracy of the model, especially at the particle-electrolyte interface, a 1-dimensional particle model was developed to evaluate single particle and dual particle suspension systems and to quantify the effect of intrinsic and extrinsic parameters on performance. The model provided information on position and bending of conduction and valence bands in the particle and at the interface to the electrolyte, where charge transport in the solution was also considered. A membrane is usually used in PC reactor designs to allow product separation which adds mechanical challenges and performance losses. To tackle this challenge, a novel reactor concept, a droplet-based photocatalytic water splitting reactor, was proposed and implemented, with inherent product separation and much lower illumination residence time of photocatalysts. Droplet emulsions were made with perflourinated oils, electrolyte and photocatalysts. The high oxygen solubility of the oil allowed for inherent trapping of the oxygen while immediately releasing hydrogen. A technoeconomic analysis is also presented which provides insight into the current state and competitiveness of levelized cost of hydrogen for PC and PEC technologies and their coupling with an alternative storage media.