Purpose: To study the prognostic value of plasmablastic morphology after autologous stem-cell transplantation for relapsed or primary refractory myeloma. Patients and Methods: Seventy-five patients were studied, Investigators blinded to the clinical derails of the individual cases reviewed bone marrow aspirate slides to determine plasmablastic classification. Plasma-blasts were defined using strict, well-described criteria. Plasmablastic morphology was considered to be present (plasmablastic myeloma) when 2% or more plasma-blasts were present in the plasma-cell population. Results: Patients underwent transplantation 5 to 88 months (median, 20 months) after the initial diagnosis of myeloma. Twenty-eight percent of patients had plasmablastic morphology. A significantly greater proportion of patients with plasmablastic morphology had abnormal cytogenetics compared with those with non-plasmablastic classification (73% v31%, respectively; P =.003). The overall survival rate measured from the time of transplantation was significantly worse in patients with plasmablastic morphology compared with those without (median survival time, 5 months v 24 months, respectively; P <.001). Progression-free survival time was shortened also, with ct median time of 4 months compared with 12 months, respectively (P <.001). In the multivariate analysis, plasmablastic classification was the most powerful prognostic factor after transplantation for both overall (P =.001)and progression-free survival rates (P <.001). We also identified three risk groups based on plasmablastic morphology: plasma-cell labeling index, lactate dehydrogenase, and cytogenetics. The median overall survival time was 38 months when none of these factors was abnormal, 17 months with one abnormal factor, and 8 months with two or more abnormal factors (P <.001), Conclusion: Plasmablastic morphology is a powerful independent predictor of poor survival rate after autologous stem-cell transplantation for relapsed or primary refractory myeloma. J Clin Oncol 17:1551-1557, (C) 1999 by American Society of Clinical Oncology.