I study the rates of, selection into, and sorting of European immigrants' secondary migration within the United States and their geographic assimilation during the Age of Mass Migration. These phenomena are recognized as important components of the economics of immigration, but data constraints have limited prior study of them in this context. As part of the debate over immigrant distribution, they were also major issues in the broader twentieth-century immigration policy debate, which were influenced by the widely held view that immigrants in the early twentieth century were less geographically mobile and specifically more attached to urban areas than were natives and earlier immigrants. I find that immigrants throughout the Age of Mass Migration were at least as likely as natives to make inter-county moves, were more attached to urban areas, were more likely to move to urban destinations, and shared natives' increasing attachment to urban areas over time. In spite of their mobility, immigrants experienced relatively little assimilation in their place-of-residence distributions relative to natives with time in the United States, though they did experience somewhat more convergence on natives in terms of urbanization. These results help to better understand immigrant assimilation and the effects of immigration during the Age of Mass Migration and imply that the contemporary views of immigrant immobility were either false, oversimplified, or the product of changes in the US economy.