Recent fusion breeder work and how breeding can be an early application of fusion R&D are reviewed. Fusion breeders are fusion reactors designed specifically to produce fissile fuel for fission reactors such as the light water reactor (LWR). Two kinds of fusion breeders are reviewed. The first uses a blanket designed to multiply neutrons by fissioning the abundant isotopes of 238U and 232Th. This design is predicted to produce enough fissile fuel for four or more LWRs and produces so much energy in the blanket that fusion performance can be reduced to a level technologically feasible within the next 10 to 15 yr. The second kind of fusion breeder uses a blanket designed to suppress fission, which enhances safety by the nonfissioning multiplication of neutrons in beryllium. This fission-suppressed fusion breeder is predicted to produce enough fissile fuel for ten or more L WRs of equal thermal power. Either kind of fusion breeder has the potential to provide a source of reasonably priced fissile fuel after the low-cost natural uranium fuel supply is gone. Thus, rapid expansion of conventional nuclear power could be provided, if necessary, to meet our nearer term needs, while at the same time providing an early application of nuclear fusion that could accelerate the commercial development of a fusion electricity generation technology to follow. Deployment scenarios show that the suppressed-fission-type fusion breeder could enable conventional nuclear plants to be expanded to 50% of the U.S. electrical capacity by the year 2050, if necessary. Despite the high development risk associated with fusion technologies, it appears that the potential advantages of the fusion breeder could be great enough to warrant an increase in research effort to the level required to determine its feasibility for commercial application and to ensure its availability when needed, provided that there is clear evidence of an increase in U.S. demand for fission power, as evidenced by new reactor orders.