AbstractWe review the design of in-vessel components for ITER, specifically the vacuum vessel, divertor, first wall and blanket-shield, and the r.f. coupling systems. The vacuum vessel is the most straightforward in concept but also the most critical since it forms the primary containment boundary. The design is a double-wall structure supported by radial plates and filled with steel balls which are directly water cooled. The structure is typically 40–70 cm thick and the shielding performance of the vessel together with the blanket-shield is sufficient to reduce the nuclear heating in the magnets to less than 8 kW. The vessel is designed to withstand an over-pressure of 2 MPa. The nominal fusion power is specified to be 1500 MW; the first wall and the divertor are each being designed to remove 80% of the corresponding a power. The conventional high recycling approach to the divertor leads to power densities two to four times higher than can be removed by practical designs for the divertor targets. However, many divertor experiments have shown that partial or full detachment of the divertor plasma from the target plates is possible in high density regimes. In such cases, the peak power density can be reduced by a factor of 5 or more, and the bulk of the power is then radiated in the divertor channel, near the X point or in the scrape-off layer. Thus the first wall and divertor are being designed to accommodate any of these possibilities. Edge localized modes and disruptions can burn through the protective plasma-gas cushion in front of the target plates, and erosion remains a major concern. While beryllium has been selected as the first-wall material, the selective use of tungsten in areas of intense charge exchange flux is under consideration. Earlier design efforts were concentrated on an integrated first wall and blanket-shield, but recent emphasis has been on an approach with a separately cooled first wall. Heating and current drive for ITER could be furnished by ion cyclotron r.f. (ICRF) or electron cyclotron r.f. (ECRF) heating, or by neutral-beam injection (NBI). All these options are being considered. ICRF is the option at present favored in view of its well-developed technology. However, the antennae become part of the first wall and their design must address reliability and maintenance issues as well as the effects of high thermal and disruption loads. ECRF and NBI are more attractive from this point of view, but their feasibility hinges on the development of suitable sources.