Costs of migration, in terms of time, energy, and mortality risk, have a strong theoretical and empirical foundation in the study of birds. We expect these costs to be most severe for extreme long‐distance migratory landbirds, whose demanding annual routines (e.g. non‐stop flights > 8000 km and return journeys > 30 000 km) may approach their maximum physiological capabilities. To explore whether this is true, we review evidence in long‐jump migratory shorebirds (Scolopacidae), focusing most on the prototypical example, the Alaska‐breeding bar‐tailed godwit Limosa lapponica baueri. Contrary to expectations, these and similar birds demonstrate high adult survival, little evidence for elevated mortality during migration, no apparent minimisation of non‐stop flight distances, and low inter‐ and intra‐individual variation in migration performance. Two key aspects of extreme migrants may explain these findings: 1) a counter‐intuitively conservative annual‐cycle strategy, which minimises risks and enables dissipation of carry‐over effects before fitness consequences arise; and 2) selection pressure during early life, which quickly removes low‐performing individuals from the population. We hypothesise that these two factors, applicable to extreme strategies in a wide range of taxa, act to truncate the range of individual quality in a population, and decrease the prevalence and detectability of carry‐over effects. Testing these hypotheses is challenging, as it requires comparative studies of demography and individual quality spectra along a continuum of extremeness. However, it has important potential implications for interpreting individual variation, designing studies of cross‐seasonal interactions or costs of migration, and recognising early‐warning signs of population decline. For example, the most extreme shorebird migrations rely on abundant but difficult‐to‐access resources; the high minimum individual performance required for survival predicts that degradation of these resource hot‐spots will propel rapid population collapse, rather than incremental declines in condition or performance. Therefore, in extreme migrants, we may paradoxically view populations as operating close to the edge, even while individuals are not.