A long-standing debate is concerned over how long premigration socioeconomic differences persisted for immigrants and their descendants who entered at the turn-of-thetwentieth century. Some researchers argue that differences exist today, over 100 years after first arrival, while others argue that most differences disappeared after the third generation. However, none of this research has directly measured pre-migration socioeconomic status nor has it directly linked immigrants to their children. I create a new panel dataset that follows immigrants and their children from the sending country through settlement. Specifically, I link ship manifest records to census records to track how long premigration socioeconomic differences persist across generations. Passenger records provide a wealth of information of individuals including the occupation before arrival. I analyze how long premigration differences persist within and between groups. Whereas premigration socioeconomic status is associated with the first generation's economic outcomes after settlement, many of these differences disappear by the second generation. These results suggest that background is not destiny for immigrant descendants. As scholars and politicians debate about whether countries should admit primarily high-skilled or low-skilled immigrants, the results from this article tell us whether such selection policies are necessary to ensure strong migrants' performance in a period of open borders.Acknowledgements: Dylan Connor, Leah Boustan, Zach Ward, and participants at McGill's Centre on Population Dynamics speaker series provided useful comments on this article. All mistakes are my own.From its inception to the mid-1920s, the United States had an open immigration policy with a few ethno-racial exceptions. Millions of immigrants entered the US and were subsequently absorbed into the labor market. A long-standing debate is concerned over how long initial differences in socioeconomic status persisted for these immigrant groups (Waters 1990; Waters and Lieberson 1989;Alba et al. 2001;Borjas 1994Borjas , 2001. Whereas some researchers argue that the United States is a melting-pot, quickly dissolving background differences, others argue that there was strong socioeconomic persistence for groups leading to long-term differences that lasted for over 100 years. Much of this research, however, has relied on synthetic-cohort analyses since there are few datasets that link immigrant parents to their offspring. However, with the development of panel datasets using full-count censuses and other historical sources, there has been a resurgence of research seeking to understand how long socioeconomic differences persist (Abramitzky et al. 2014;Connor 2018b;Ward 2018). This study enters this debate by investigating how long differences in premigration socioeconomic status lasted within and across immigrant groups who entered at the turn of the twentieth century.Specifically, this article develops a new dataset that tracks within-family intergenerational mobility ...