Immunophenotypic investigation of minimal residual disease (MRD) has traditionally been based on the investigation of phenotypic aberrants at diagnosis to be used later as a target for MRD detection. This approach has several shortcomings (it is only applicable to patients with aberrant phenotypes, requires a diagnostic sample, and is patient-specific) and therefore a search for simpler alternatives is warranted. The present study is based on the hypothesis that in precursor-B-ALL patients the persistence of residual leukaemic cells may induce abnormalities in the precursor-B-cell compartment in bone marrow (BM) and these could be used as a criteria to predict relapse. These abnormalities may include: (1) the presence of an increase in the frequencies of immature B cells (CD34+/CD19+ or CD20-/CD19+) or (2) the existence of an altered B-cell differentiation pathway due to a blockade or to the presence of B cells outside the normal pathway. A total of 180 BM samples from 45 consecutive precursor-B-ALL patients who achieved morphological complete remission (CR) were analysed by multiparametric flow cytometry. Our results show that a significant increase in immature B-cell subsets or an altered B-cell differentiation predicts a high relapse rate (P<0.01) and a shorter disease-free survival (P<0.01). Moreover, abnormalities in either of these two criteria detected at specific time points during follow-up (end of induction, maintenance, or after treatment) were associated with a significantly shorter disease-free survival (P<0.01). In summary, the investigation of abnormalities in B-cell differentiation is a relatively simple and cheap approach for predicting relapse in precursor-B-ALL patients.