Mobile species are particularly affected by artificial barriers requiring large investments to restore connectivity. However, few large‐scale, long‐term studies have investigated the ecological outcomes of restoring connectivity for these species. Our study, spanning 15–20 years, quantified response trajectories, which represent temporal trends following disturbance, of three native salmonids colonizing 20 km of protected habitat following restoration of fish passage at Landsburg Dam, Cedar River, WA, in 2003. Built in 1901, the dam blocked the upriver movement of native anadromous coho and Chinook salmon and nonanadromous mountain whitefish for 102 years. Restoration effectiveness was also assessed by comparing temporal trends in freshwater productivity of juvenile coho and Chinook salmon in the Cedar River after restoration to a nearby undammed subbasin. We also compared summer densities of juvenile coho and Chinook salmon, and mountain whitefish above the dam measured a decade after restoration to undammed reference systems. Anadromous salmon and nonanadromous mountain whitefish populations increased linearly or nonlinearly following restoration. The positive, asymptotic response represented by adult Chinook salmon counts indicates a slowing in population recovery rate, plateauing a decade after restoration. In contrast, annual abundance of adult coho salmon increased at a constant rate, indicating additional capacity 15 years post‐restoration. Salmonid compositional diversity, driven largely by juvenile coho salmon, also increased nonlinearly, plateauing in a decade. We observed substantial spatial variation in the temporal response, as juvenile coho salmon and mountain whitefish population expansion slowed linearly with upstream distance from the restoration site. There was evidence that some of the annual variation in salmonid biomass in summer was a result of discharge variability in winter and spring, with biomass declining as flow variability increased. Species reintroduction and establishment had no discernible effect on stream‐rearing salmonids living above the dam before restoration, while increasing coho freshwater productivity at the subbasin scale. Results from our study showed recolonization by three mobile species, each with a unique life history, takes at least a decade or more and was dependent on species and life stage, size of the spawning population, distance from restoration site, and annual variability in streamflow.