The detection of residual molecular and cytogenetic disease was prospectively compared in patients with Philadelphia-chromosome (Ph1) positive first chronic phase chronic myelogenous leukemia (CML) who underwent allogeneic transplantation of unmanipulated peripheral blood stem cells (PBSCT) (n = 29) or bone marrow (BM) (n = 62) using genotypically HLA-identical sibling donors or partially HLA-matched extended family donors. A molecular relapse (MR), as defined by two consecutive positive polymerase chain reaction (PCR) assays for the detection of M-bcr-abl transcripts in a 4-week interval, was found in two of 29 (7%) patients after PBSCT compared with 20 of 62 (32%) patients after bone marrow transplantation (BMT). This corresponds to a 4-year molecular relapse estimate (+/- standard error) of 7% +/- 5% after PBSCT and of 44% +/- 8% after BMT (P <.009). With identical follow-up periods of survivors in both patient subsets between 6 and 55 months (median, 28 months), 14 of the 20 patients with MR after BMT progressed to an isolated cytogenetic (n = 10) or a hematologic (n = 4) disease recurrence, resulting in a 4-year cytogenetic relapse estimate of 47% +/- 11%, while none of the patients after PBSCT has so far relapsed (P <.006). Multivariate analysis including all potential influencial factors of posttransplant disease recurrence identified the source of stem cells (P <.02) as the only independent predictor of molecular relapse. In conclusion, this prospective comparison of molecular and cytogenetic residual disease demonstrates that peripheral blood stem cell transplants have a more pronounced activity against residual CML cells than bone marrow transplants. Prospective randomized trials comparing PBSCT and BMT in patients with first chronic phase Ph1-positive CML are strictly required to further substantiate differences in the antileukemic activity of the two stem cell sources.