Magnetic confinement fusion (MCF) is an approach to generate thermonuclear fusion power that uses magnetic fields to confine fusion fuel in the form of a plasma. Magnetic confinement is one of two major branches of controlled fusion research, along with inertial confinement fusion.

Fusion reactions for reactors usually combine light atomic nuclei of deuterium and tritium to form an alpha particle (helium-4 nucleus) and a neutron, where the energy is released in the form of the kinetic energy of the reaction products. In order to overcome the electrostatic repulsion between the nuclei, the fuel must have a temperature of hundreds of millions of kelvin, at which the fuel is fully ionized and becomes a plasma. In addition, the plasma must be at a sufficient density, and the energy must remain in the reacting region for a sufficient time, as specified by the Lawson criterion (triple product). The high temperature of a fusion plasma precludes the use of material vessels for direct containment. Magnetic confinement fusion attempts to use the physics of charged particle motion to contain the plasma particles by applying strong magnetic fields.

Tokamaks and stellarators are the two leading MCF device candidates as of today. Investigation of using various magnetic configurations to confine fusion plasma began in the 1950s. Early simple mirror and toroidal machines showed disappointing results of low confinement. After the declassification of fusion research by the United States, United Kingdom and Soviet Union in 1958, a breakthrough on toroidal devices was reported by the Kurchatov Institute. Its tokamak demonstrated a temperature over around 10 million kelvin and milliseconds of confinement time, and was confirmed by a visiting British team. Since then, tokamaks became the dominant line of research globally with JET, TFTR and JT-60 being constructed and operated. The ITER tokamak experiment under construction, which aims to demonstrate scientific breakeven, will be the world's largest MCF device.

Experiments with deuterium-tritium plasmas in TFTR created 1.6 GJ fusion energy during 1993-1996. The peak fusion power was 10.3 MW from 3.7 x 10¹⁸ reactions per second, and peak fusion energy created in one discharge was 7.6 MJ. Subsequent experiments in JET achieved a peak fusion power of 16 MW (5.8 x 10¹⁸ /s) during DTE1 , and a discharge producing 69 MJ during the recent DTE3 (consuming 0.2 mgm of D and T).

The current record of fusion power generated by MCF devices is held by JET. In 2021, JET sustained a gain factor of Q = 0.33 for 5 seconds and produced 59 megajoules of energy. The record for the gain factor was 0.67 achieved in 1997, and the gain in the plasma core Q_core for that discharge is calculated to be 1.3, using the TRANSP integrated modeling code.

While early stellarators of low confinement in the 1950s were overshadowed by the initial success of tokamaks, interests in stellarators re-emerged owing to their inherent capability for steady-state and disruption-free operation, distinct from tokamaks. The world's largest stellarator experiment, Wendelstein 7-X, began operation in 2015.

One of the challenges of MCF research is the development and extrapolation of plasma scenarios to power plant conditions, where good fusion performance and energy confinement must be maintained. Potential solutions to other problems such as divertor power exhaust, mitigation of transients (disruptions, runaway electrons, edge-localized modes), handling of neutron flux, tritium breeding and the physics of burning plasmas are being actively studied. Development of new technologies in plasma diagnostics, real-time control, plasma-facing materials, high-power microwave sources, vacuum engineering, cryogenics and superconducting magnets are essential in MCF research.

A major area of research in the early years of fusion energy research was the magnetic mirror. Most early mirror devices attempted to confine plasma near the focus of a non-planar magnetic field generated in a solenoid, with the field strength increased at either end of the tube. In order to escape the confinement area, nuclei had to enter a small annular area near each magnet. It was known that nuclei would escape through this area, but by adding and heating fuel continually it was believed this could be overcome.