Amazonia now contains vast areas of secondary forest because of widespread regeneration following timber harvests, yet the value of secondary forest to wildlife remains poorly understood. Secondary forest becomes structurally similar to primary forest after abandonment, and therefore we predicted that avian movement across the interface of primary and secondary forest (hereafter “the interface”) would gradually increase with time since abandonment until recovery to pre-isolation levels. From 1992 to 2011, we captured 2,773 understory birds of 10 foraging guilds along the interface of primary forest fragments and zero- to 30-year-old secondary forest at the Biological Dynamics of Forest Fragments Project near Manaus, Brazil. Our objectives were to understand the differences in land-use history that affect cross-interface movement and to determine how long it takes each guild to recover to pre-isolation capture rates. Across guilds, age of secondary forest within 100 m of the interface was the most important explanatory variable affecting capture rates; rates increased with age of secondary forest for all guilds except non-forest species. Mean recovery to pre-isolation was 26 years (asymmetric SE = 13 years below and 16 years above estimate) after secondary forest abandonment and 9 of 10 guilds recovered within 13 to 34 years. In the slowest guild to recover, terrestrial insectivores, 6 of 12 species were never caught along the interface, and we projected that this guild would recover in ∼60 years. Our recovery estimates quantify the dynamic permeability of the interface and contribute to a better understanding of the value of secondary forests as corridors among primary forest fragments.