The current treatment of acute promyelocytic leukemia (APL, also called AML3 subtype) is focused on differentiating agents such as the vitamin A derivative all-trans retinoic acid (ATRA). This agent is a novel and very promising therapy for this disease characterized cytogenetically by a translocation t(15;17)(q21;q22) involving the alpha retinoic acid receptor on chromosome 17 and the PML gene on chromosome 15. Clinical trials have demonstrated that ATRA followed by or combined with conventional chemotherapy may be more beneficial than chemotherapy for inducing complete remission. Unfortunately, ATRA as a single agent, does not appear able to maintain patients in remission (median 5 months), and when relapse occurs resistance to a second induction of ATRA therapy is observed in almost all cases. Recently our laboratory investigated whether specific features of the AML3 cells at relapse could explain the in vivo resistance observed. We have demonstrated that AML3 patients' cells (from four patients) at relapse show high levels of CRABP, a cytosolic retinoic acid binding protein and this protein was not detected prior to ATRA therapy. Relapse-AML3 cells (n = 12) showed reduced differentiation induction when compared with 'virgin'-AML3 cells. Results from this study suggest that CRABP could modulate ATRA cellular concentrations reaching the nucleus. This induced ATRA hypercatabolytic state should be monitored during consolidation therapy and at relapse by evaluating CRABP and RA metabolite levels, in order to detect ATRA resistance in patients with AML3.