The terahertz frequency range, often referred to as the “terahertz” gap, lies wedged between microwave at the lower end and infrared at the higher end of the spectrum, occupying frequencies between 0.3 and 3.0 THz. For a long time, applications in THz frequencies had been limited to astronomy and chemical sciences, but with advancements in THz technology in recent years, it has shown great promise in a wide range of applications ranging from security screening for concealed weapons and contraband detection to global environmental monitoring, nondestructive quality control, and ultra-fast wireless communication. Up until recently, the terahertz frequency range has been mostly addressed by high-mobility compound III-V processes, expensive nonlinear optics, or cryogenically cooled quantum cascade lasers. A low-cost, room temperature alternative can enable the development of such a wide array of applications, not currently accessible due to cost and size limitations. In this thesis, we will discuss our approach towards development of integrated terahertz technology in silicon-based processes. In the spirit of academic research, we will address frequencies close to 0.3 THz as “terahertz.” In this chapter, we address both fronts of integrated THz systems in silicon: THz power generation, radiation and transmitter systems, and THz signal detection and receiver systems. THz power generation in silicon-based integrated circuit technology is challenging due to lower carrier mobility, lower cut-off frequencies compared to compound III-V processes, lower breakdown voltages, and lossy passives. Radiation from silicon chips is also challenging due to lossy substrates and the high dielectric constant of silicon. In this chapter, we will describe novel ways of combining circuit and electromagnetic techniques in a holistic design approach, which can overcome limitations of conventional block-by-block or partitioned design methodology, in order to generate high-frequency signals above the classical definition of cut-off frequencies (ft/fmax). We demonstrate this design philosophy in an active electromagnetic structure, which we call distributed active radiator. It is inspired by an inverse electromagnetic (EM) approach, where instead of using classical circuit and electromagnetic blocks to generate and radiate THz frequencies, we formulate surface (metal) currents in silicon chips for a desired THz field profile and develop active means of controlling different harmonic currents to perform signal generation, frequency multiplication, radiation and lossless filtering, simultaneously in a compact footprint. By removing the artificial boundaries between circuits, electromagnetics, and antenna, we open ourselves to a broader design space. This enabled us to demonstrate the first 1-mW effective-isotropic-radiated-power (EIRP) THz (0.29 THz) source in CMOS with total radiated power being three orders of magnitude more than previously demonstrated. We also proposed a near-field synchronization mechanism, which is a scalable method of realizing large arrays of synchronized autonomous radiating sources in silicon. We also demonstrate the first THz CMOS array with digitally controlled beam-scanning in 2D space with radiated output EIRP of nearly 10 mW at 0.28 THz.