Since 1992, efforts to recover the US federally endangered Great Lakes piping plover (Charadrius melodus) have included population supplementation with captive-reared young raised from abandoned eggs. Using banding data collected 1993–2008 and Cormack–Jolly–Seber models in program MARK we estimated resighting rates (ρ) and apparent survival (Φ) of first-year (1) and after-first-year (2+) plovers. Reproductive success was measured by estimating mean number of eggs laid, chicks hatched and chicks fledged per wild-reared or captive-reared parent, and these values were compared with a permutation test. The best-supported mark-resighting model indicated wild-reared plovers had higher Φ1 and Φ2+ than captive-reared plovers. Breeding propensity influenced detection of wild plovers, whereas unique band combinations facilitated detection of yearling captive-reared plovers. Captive-reared (n = 10) and wild-reared (n = 57) plovers laid similar numbers of eggs, but wild-reared plovers hatched 36% more chicks and fledged 56% more young. Reproductive values derived from matrix models suggest captive-reared piping plovers are less fit than similarly aged wild-reared birds upon release and demonstrate reduced fitness in subsequent years. The Great Lakes captive-rearing effort has successfully produced a minimum of 10 breeding adults from 192 eggs that otherwise would have had no reproductive value; these captive-reared individuals now constitute up to 3% of the total population. Small scale salvage and captive-rearing of abandoned eggs should be considered as a technique to supplement existing recovery efforts in highly imperiled populations.