We tested whether captive breeding and release is a feasible restoration strategy for the Harpy Eagle (Harpia harpyja) where suitable unoccupied habitat remains within its former range. From 1987 through 2006, 18 Harpy Eagles participated in a captive breeding program started in Boise, Idaho, and continued in Panama from 2001. From 131 eggs laid, 44 eagles were fledged. Most young were produced by just three females in the program, and at a higher annual rate after the birds were moved from Boise to Panama. Re-laying induced by collecting eggs for artificial incubation increased the number of viable eggs laid per female each breeding season up to six, but may have reduced female reproductive lifetime. Including rehabilitated eagles hatched in the wild, we released 49 eagles from 1998 through 2009. When the last released eagle with a functioning radio transmitter died in 2011, 63% were known or presumed to be dead, 31% were missing and possibly alive, and 6% were back in captivity. Shooting (44%) was the primary cause of death. Behavior interpreted as aggression toward humans was sufficiently frequent (23% of released eagles) in captive-raised and wild-rehabilitated eagles after release to be a concern for public safety and a potential cause of shooting deaths. This study demonstrated that it is feasible to breed Harpy Eagles in captivity at high rates needed for species restoration. It is possible to release captive-reared and rehabilitated Harpy Eagles to the wild, and is most cost effective (i.e., resulting in the highest survival to hunting-independence) when eagles are released close to the age of independence. Preventing shooting and other kinds of human persecution, and protecting remaining forest habitat, are the most urgent conservation needs for the Harpy Eagle.