III-V semiconductor nanowires have unique properties that make them ideal for advanced photodetectors on inexpensive substrates. For example, they exhibit enhanced or polarization-dependent light absorption, they can form complex heterostructures, and their charge carrier lifetimes can be engineered. Research to-date has focused on vertical nanowires which are grown perpendicular to the growth substrate, but fabricating devices out of such nanowires is difficult and time-consuming. They can either be left in a vertical configuration, an approach which is technically challenging and vulnerable to performance losses due to disorder, or they can be transferred to a secondary substrate, which requires tailoring individual devices one-by-one, a process which is inherently not scalable. On the other hand, nanowires which are grown directly in a horizontal configuration, parallel to the growth substrate, can be made in wafer-scale arrays and networks, with facile alignment to subsequent device layers. In this thesis, we investigate scalable approaches to the epitaxial growth of horizontal III-V nanowires and related nanostructures for photodetection applications. We begin by introducing key concepts related to epitaxy, photodetection, and semiconductor nanostructure characterization, before presenting the main results from my PhD thesis work. We characterize and model the nanoscale growth phenomena which govern the formation of GaAs nanoridges grown by selective area metalorganic vapor-phase epitaxy (MOVPE) on GaAs substrates. Next, we show how to functionalize such nanostructures, using them as a template for subsequent InAs nanowire growth. We demonstrate various ways to tune the properties of the nanowires and identify key differences in the morphology of similar structures grown by molecular beam epitaxy (MBE). We then fabricate THz photoconductive detectors from MBE-grown InAs nanowires and characterize their THz response using IR optical pumping across the full range of telecommunications wavelengths. We correlate a transition from direct to integrating-type detection to the morphology of the nanowires, attributing the effect to different recombination velocities on different crystal faces. Next, we explore two routes to integrating III-V nanostructures onto Si substrates. We firstly report the selective area growth of GaAs nanoridges on Si (100) substrates by both MBE and MOVPE, showing the evolution from the nucleation stage to the final nanostructure. We identify how some, but not all, planar defects can be effectively trapped inside of the patterned trenches, and discuss specific challenges related to reproducibility for each kind of epitaxial growth system. Then we turn our attention to a less conventional approach to epitaxy: rapid melt growth. We process MBE-deposited polycrystalline thin films of InSb into monocrystalline nanowire-like structures and characterize their optoelectronic properties for IR detection. Finally, we conclude by summarizing the key results and giving an outlook for future research directions. Overall, I believe the results presented here represent important advances in transforming the theoretical advantages of III-V nanowires into realizable devices that can be implemented via scalable processes.