Sixty precursor B-cell acute lymphoblastic leukemia (ALL) patients were analyzed for the configuration of their immunoglobulin (Ig) genes. Rearrangements and/or deletions of the Ig heavy chain (IgH), Ig kappa chain (Ig kappa), and Ig lambda chain (Ig lambda) genes were detected in 98, 48, and 23% of cases, respectively. Although these percentages suggest the presence of a hierarchical order in IgH and Ig light chain (IgL) gene rearrangements during B-cell differentiation, no correlation was found between the immunophenotype of the precursor B-ALL and the arrangement patterns of their IgH and IgL genes. Multiple rearranged IgH gene bands, generally differing in density, were found in 27 (45%) of the precursor B-ALL in various restriction enzyme digests. Cytogenetic data were used to determine whether the presence of more than two rearranged IgH gene bands was caused by hyperdiploidy of chromosome 14 or other chromosome 14 aberrations. The combined cytogenetic and IgH gene data allowed the precursor B-ALL to be divided into three groups: a monoclonal group (n = 36; 60%), a biclonal group (n = 16; 27%), and an oligoclonal group (n = 8; 13%). In five biclonal ALL biclonality at the Ig kappa gene level was also found. Such subclone formation was not detected at the Ig lambda gene level. As the detection limit of the Southern blot technique is 2-5%, it might well be that small subclones remained undetected, implying that the frequency of subclone formation at the IgH gene level in precursor B-ALL is probably higher than 40%. It has been suggested that precursor B-ALL with multiple IgH gene rearrangements have a higher tendency to relapse. Although higher relapse rates were found in the oligoclonal group (53%) and in the combined bi-oligoclonal group (33%) compared with the monoclonal group (20%), the log rank trend test showed no significance. The occurrence of multiple subclones in precursor B-ALL as found by IgH gene analyses will severely hamper the detection of minimal residual disease using the polymerase chain reaction (PCR) mediated amplification of 'tumor-specific' IgH gene junctional regions, because it cannot be predicted which detectable (or undetectable) subclone will cause minimal residual disease and/or relapse. Therefore it can be expected that the PCR technique will frequently produce false negative results during the follow-up of precursor B-ALL.