TFTR is one of four large 2nd generation tokamak devices, the others being T15, JT60 and JET. TFTR was completed ten months ago, the others are in various stages of construction and initial operation. TFTR, as well as its contemporaries, has important contributions to make towards our understanding of plasma conditions in the thermonuclear reactor regimeò One of the main objectives of TFTR is to produce fusion power densities approaching those in a fusion reactor, 1 W cm−3, at Q ∼ 1–2. This will be done in two ways, first by utilizing the “two component torus” concept whereby high energy deuterium neutral beams at 120 keV collide with a tritium target plasma, thereby making substantial use of the direct driven fusion system and, later, by bulk plasma heating of D-T plasma with neutral beams, ohmic heating, RF, and alpha particles created by longer pulse operations with a D-T reacting plasma. TFTR will be the first tokamak device to use deuterium and tritium fuel. TFTR was designed with remote handling considerations. Table I lists the D-T pulses for Q ≅ 2. The physics objectives of TFTR dictated a relatively compact tokamak with a large diameter plasma. This basic design concept resulted in a machine constructed of massive, high strength, non-magnetic components which require installation to extremely tight tolerances. In this paper the assembly process for TFTR is reviewed from start of assembly in January 1982 to the scheduled completion in December 1982. The techniques, tooling and procedures necessary to complete the tokamak on this aggressive schedule are described. Work-arounds and interim assemblies, used to maintain forward momentum on main assembly, will be discussed. QA, industrial engineering approaches, as-built and other fabrication-related problems, their solutions, and their effect on the design, are summarized. Preliminary start-up performance data for the initial ohmic heating phase will also be discussed. Since TFTR is a major component of the U.S. fusion program and is designed to handle D-T, implications of its experimental program performance will be summarized. Some comments on the remote handling program will also be presented. All of the work reported on was supported by the Office of Fusion Energy of the U.S. Department of Energy.