The sensitivity and clinical utility of the polymerase chain reaction (PCR) assay for the detection of BCR-ABL gene rearrangement was compared to conventional cytogenetics for the Philadelphia chromosome (Ph1) in adult acute lymphoblastic leukemia (ALL) patients entered onto a single clinical trial. Ninety-three patients had evaluable PCR assays for both the p190bcr-abl and p210bcr-abl type of BCR-ABL gene rearrangements. Twenty-one of 93 patients (23%) were positive for the BCR-ABL rearrangement by the PCR assay. Fourteen of these patients had the p210brc-abl BCR-ABL rearrangement characteristically seen in CML patients, while seven had the p190bcr-abl rearrangement seen in ALL alone. Of 61 patients analyzed, both with conventional cytogenetics and PCR, eight (13%) were positive for the Ph1, while 14 (23%) were positive for the BCR-ABL rearrangement by the PCR assay. Discordance between the PCR assay and cytogenetics occurred in eight cases where the PCR assay was positive and the cytogenetics negative, and two cases where the PCR assay was negative and cytogenetics positive. PCR positivity did not correlate with treatment response, survival, or relapse-free survival, but there was a higher percentage of L2 FAB morphology in the PCR+ cases compared to the PCR-cases (67 vs. 28%, p = 0.003). In addition, the data suggested that patients with a p190bcr-abl rearrangement have a better response to induction therapy, but a worse relapse-free survival compared to patients with a p210bcr-abl breakpoint, but these differences were not statistically significant. These data suggest that PCR and conventional cytogenetics may provide complementary information, since there appear to be a subset of patients who are Ph1-negative yet BCR-ABL positive by PCR. Further studies will be required to determine the prognostic significance of the detailed information about BCR-ABL breakpoints that is available from the PCR assay.