An option in the long-duration exploration of space, whether on the Moon or Mars or in a spacecraft on its way to Mars or the asteroids, is to utilize a bioregenerative life-support system in addition to the physicochemical systems that will always be necessary. Green plants can use the energy of light to remove carbon dioxide from the atmosphere and add oxygen to it while at the same time synthesizing food for the space travelers. The water that crop plants transpire can be condensed in pure form, contributing to the water purification system. An added bonus is that green plants provide a familiar environment for humans far from their home planet. The down side is that such a bioregenerative life-support system--called a controlled environment life-support system (CELSS) in this paper--must be highly complex and relatively massive to maintain a proper composition of the atmosphere while also providing food. Thus, launch costs will be high. Except for resupply and removal of nonrecycleable substances, such a system is nearly closed with respect to matter but open with respect to energy. Although a CELSS facility is small compared to the Earth's biosphere, it must be large enough to feed humans and provide a suitable atmosphere for them. A functioning CELSS can only be created with the help of today's advanced technology, especially computerized controls. Needed are energy for light, possibly from a nuclear power plant, and equipment to provide a suitable environment for plant growth, including a way to supply plants with the necessary mineral nutrients. All this constitutes the biomass production unit. There must also be food preparation facilities and a means to recycle or dispose of waste materials and there must be control equipment to keep the facility running. Humans are part of the system as well as plants and possibly animals. Human brain power will often be needed to keep the system functional in spite of the best computer-driven controls. The particulars of a CELSS facility depend strongly on where it is to be located. The presence of gravity on the Moon and Mars simplifies the design for a facility on those bodies, but a spacecraft in microgravity is a much more challenging environment. One problem is that plants, which are very sensitive to gravity, might not grow and produce food in the virtual absence of gravity. However, the experience with growing super-dwarf wheat in the Russian space station Mir, while not entirely successful because of the sterile wheat heads, was highly encouraging. The plants grew well for 123 days, producing more biomass than had been produced in space before. This was due to the high photon flux available to the plants and the careful control of substrate moisture. The sterile heads were probably due to the failure to remove the gaseous plant hormone, ethylene, from the Mir atmosphere. Since ethylene can easily be removed, it should be possible to grow wheat and other crops in microgravity with the production of viable seeds. On the ground Biosphere-2 taught us several lessons about the design and construction of a CELSS facility, but Bios-3 came much closer to achieving the goals of such a facility. Although stability was never completely reached, Bios-3 was much more stable than Biosphere-2 apparently because every effort was made to keep the system simple and to use the best technology available to maintain control. Wastes were not recycled in Bios-3 except for urine, and inedible plant materials were incinerated to restore CO2 to the atmosphere. Since much meat (about 20% of calories) was imported, closure in the Bios-3 experiments was well below 100%. But then, a practical CELSS on the Moon might also depend on regular resupply from Earth. Several important lessons have been learned from the CELSS research described in this review.