The first such reactor to generate electricity was the US Experimental Breeder Reactor 1 (EBR 1). This started in 1951 with a capacity of 200 kWe. It was fuelled by highly enriched uranium-235. In common with future fast reactor designs, the core was small and compact. The fuel pins were just 1.25 cm in diameter. The core consisted of 217 pins in a hexagonal lattice. The coolant was a sodium/potassium alloy, surrounding the central region was a blanket region containing rods of natural uranium. EBR 1 operated until 1963 and yielded considerable information on liquid metal fast breeder reactor (LMFBR) technology. A second reactor EBR 2, 15.7 MW, was also built on the Arco site in Idaho.

A 60 MW commercial reactor, Enrico Fermi 1 went critical in 1963. This reactor underwent a serious loss of coolant accident in 1966. It restarted for a few years but was finally shut down in 1970.

The US fast reactor programme continued with various test facilities until 1983, e. g. the southwest experimental fast oxide reactor (SEFOR) at Arkansas, the transient reactor test experiment (TREAT) at Argonne and the fast flux test facility (FFTF) at Hanford.

Within Europe, the United Kingdom atomic energy authority (UKAEA) built several research reactors before the Dounreay fast reactor (DFR) was commissioned and became critical in 1959. DFR had a modest electrical capacity of 14 MWe. It was closed down in 1977. The prototype fast reactor (PFR) had an electrical output of 254 MWe and entered service in 1975. It operated for over a decade before being shut down.

This sodium-cooled fast reactor was a pool type design. A pool of sodium is contained in a vessel with sodium pumped through the core by pumps contained within the pool. The hot sodium then passes through an intermediate heat exchanger; transferring heat to a second sodium-cooled loop. The latter transfers heat to a water/steam loop via the steam generator. This tertiary loop system ensures that any radionuclides produced in the primary vessel remain in the vessel and are not transferred to the steam generator.

In this type of reactor design, the reactor functions on fast neutrons, there is no moderator.

In France, a similar 250 MW prototype was also built (Phenix), which was then followed by a commercial sized plant (Superphenix), the latter commissioned in 1986 (but now closed down permanently).

Other countries have explored the production of fast reactors, e. g. Germany, Japan, India and the former Soviet Union.

LMFBRs have a number of advantages. Liquid metals have desirable thermophysical properties. The coolant has a low melting point, coolants can be chosen, e. g. sodium and potassium, which have low neutron absorption. Sodium has a high thermal conductivity, albeit a lower specific heat than water and it has a high boiling point, etc.

LMFBRs also suffer from a number of disadvantages and problems. There are concerns over the use of sodium since it is highly reactive to oxygen and water. There is a potential problem of isolation of the sodium and water-cooling loops. There have been problems in the steam generators of fast reactors.

In recent years the development of fast reactors at the commercial scale has slowed down. Nevertheless, the potential for fast reactors exists and is still under review in some countries. Fast reactors are again under consideration in the US Generation IV programme.

Historically, the fast reactor has always been considered in relation to its fuel cycle, its ability to burn and breed plutonium. In addition, most reactors produce plutonium, in differing amounts, which can in principle be recovered for utilisation in a fast reactor fuel cycle. However, there are safety and economic issues associated with fuel reprocessing, these are considered later. Plutonium can also be burnt in thermal reactors to improve the economics of the thermal fuel cycle.