Using historical, longitudinal data on individuals, we track the earnings of immigrant and U.S.-born women. Following individuals, instead of synthetic cohorts, avoids biases in earnings-growth estimates caused by compositional changes in the cohorts that are followed. The historical data...