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. We find a large decline in the redefault rate of both modified and self-cured mortgages over this period, but the improvement was greatest for modifications. Our analysis has identified several important factors contributing to the greater improvement for modified loans, including an increasing share of principal-reduction modifications, which appear to be more effective than other types of modification and increasingly generous modification terms (larger payment reductions). The favorable impacts of principal and payment reductions on household finances were enhanced by improving economic conditions, resulting in more effective modifications. Even after accounting for these factors, we still observe a larger decline in the redefault rate for modifications compared with similarly situated self-cured loans. This residual effect may reflect servicer “learning-by-doing”; that is, servicers gained knowledge as modification activity ramped up, resulting in more successful modification programs for later cohorts.