PURPOSE Autologous hematopoietic cell transplantation, or autotransplantation, is effective in light-chain amyloidosis (AL), but it is associated with a high risk of early mortality (EM). In a multicenter randomized comparison against oral chemotherapy, autotransplantation was associated with 24% EM. We analyzed trends in outcomes after autologous hematopoietic cell transplantation for AL in North America. PATIENTS AND METHODS Between 1995 and 2012, 1,536 patients with AL who underwent autotransplantation at 134 centers were identified in the Center for International Blood and Marrow Transplant Research database. EM and overall survival (OS) were analyzed in three time cohorts: 1995 to 2000 (n = 140), 2001 to 2006 (n = 596), and 2007 to 2012 (n = 800). Hematologic and renal responses and factors associated with EM, relapse and/or progression, progression-free survival and OS were analyzed in more recent subgroups from 2001 to 2006 (n = 197) and from 2007 to 2012 (n = 157). RESULTS Mortality at 30 and 100 days progressively declined over successive time periods from 11% and 20%, respectively, in 1995 to 2000 to 5% and 11%, respectively, in 2001 to 2006, and to 3% and 5%, respectively, in 2007 to 2012. Correspondingly, 5-year OS improved from 55% in 1995 to 2000 to 61% in 2001 to 2006 and to 77% in 2007 to 2012. Hematologic response to transplantation improved in the latest cohort. Renal response rate was 32%. Centers performing more than four AL transplantations per year had superior survival outcomes. In the multivariable analysis, cardiac AL was associated with high EM and inferior progression-free survival and OS. Autotransplantation in 2007 to 2012 and use of higher dosages of melphalan were associated with a lowered relapse risk. A Karnofsky score less than 80 and creatinine levels 2 mg/m(2) or greater were associated with worsened OS. CONCLUSION Post-transplantation survival in AL has improved, with a dramatic reduction in early post-transplantation mortality and excellent 5-year survival. The risk-benefit ratio for autotransplantation has changed, and randomized comparison with nontransplantation approaches is again warranted.