Quantifying survival and understanding underlying sources of variation are important for developing population models and informing management decisions. We estimated apparent survival (i.e., true survival less permanent emigration) for adult female white‐winged scoters (Melanitta fusca) and lesser scaup (Aythya affinis) breeding at a northern boreal forest site in western Canada, 2002–2008. We also evaluated variation in survival relative to indices of breeding status, individual quality, spring weather conditions, local small‐mammal abundance, and overwinter climate. Breeding female scoters had higher apparent survival than did nonbreeding females, suggesting that breeders had higher survival or fidelity to the study area, or that more nonbreeders were transient birds that may have bred elsewhere in subsequent years. Apparent survival rate for breeding female scoters was unrelated to other individual and environmental covariates. Nest‐trapped female scaup had higher apparent survival rates than did prenesting females captured in decoy traps, implying that more pre‐nesters dispersed permanently or died after marking. Nesting female scaup with higher body condition or those in years when small mammals were more abundant had higher apparent survival; associations between survival and other environmental covariates were less certain. Overall, apparent survival rate of breeding adult female scoters was lower than reported for scoters from other North American locations or for other sea duck species, whereas estimates for nesting female scaup were comparable with those from other boreal and prairie‐parkland locations. Our results indicate that for scaup in this region, factors influencing female body condition, such as maintaining high‐quality habitat, areas with abundant food or low disturbance, could improve annual survival. © The Wildlife Society, 2019