This paper examines changes in the redefault rate of mortgages that were selected for modification during 2008–2011, compared with that of similarly situated self-cured mortgages during the same period. We find that while the performance of both modified and self-cured loans improved dramatically over this period, the decline in the redefault rate for modified loans was substantially larger, and we attribute this difference to a few key factors. First, the modification terms regarding repayments have become increasingly more generous, including more principal reduction, resulting in greater financial relief to the borrowers. Second, modifications in later vintages also benefited from improving economic conditions. Modifications became more effective as unemployment rates declined and home prices recovered. Third, we find that the difference between redefault rate improvement between modified loans and self-cured loans continue to persist even after controlling for all the relevant risk and economic factors. We attribute this difference to the servicers’ learning process — such as data collection and information sharing among industry participants — known as “learning-by doing.” Early in the mortgage crisis, many servicers had limited experience selecting the best borrowers for modification. As modification activity increased, lenders became more adept at screening borrowers for modification eligibility and in selecting appropriate modification terms. Our empirical findings suggest that mortgage modification effectiveness could be enhanced through the industry’s “learning-by-doing” process.