The terahertz (THz) frequency range lies between microwaves and visible radiation in the electromagnetic spectrum. This relatively unexplored region of the spectrum is attracting significant interest internationally due to its ability to discriminate samples chemically, to identify differences in physical structure, and to penetrate non-polar materials. However, it lacks a compact semiconductor source of ultrashort THz pulses, suitable for applications including ultrafast spectroscopy for materials analysis, and THz frequency comb generation with applications for trace gas sensing and atmospheric science. Although the quantum cascade laser (QCL) is a promising compact semiconductor source of THz radiation, its success in creating ultrashort pulses is limited due to the inherent fast gain recovery time. The exciting possibility has been proposed of exploiting the phenomenon of self-induced transparency (SIT) for passive laser modelocking, in which the pulse duration is mediated through the process of Rabi-flopping. The aim of the TERAULTRA project was to explore, for the first time, the feasibility of generating ultrashort THz pulses from QCLs by exploiting SIT phenomenon. In working towards this goal, this project aimed to explore, both theoretically and experimentally, the underlying phenomenon of Rabi-flopping and SIT in quantum heterostructures and QCLs in the THz frequency range. The project also aimed to explore alternative approaches to achieving broadband emission in THz QCLs such as using cascaded two-photon emissions. The TERAULTRA project: demonstrated, for the first time, SIT effects in quantum heterostructures at THz frequencies; designed, experimentally investigated and optimised THz QCL heterostructures for SIT-modelocking and short pulse generation; and proposed and designed the first two-photon emitting THz QCLs.