Most acute promyelocytic leukemia (APL) cases have t(15;17)(q22;q21) chromosomal translocation and PML-RARalpha chimeric gene which blocks granulocytic differentiation. The introduction of all-trans-retinoic acid (ATRA) and arsenic compounds, especially arsenic trioxide (As(2)O(3)), has provided good models to study not only differentiation and/or apoptosis therapy but also molecular target-based cancer treatment. In vivo and in vitro investigations have shown that both agents are able to induce differentiation of APL cells: ATRA tends to induce terminal differentiation, while low-dose As(2)O(3) can induce partial differentiation. Significant progress has been made in understanding the molecular mechanisms of APL pathogenesis and differentiation therapy. Pharmacological concentrations (0.1 approximately 1 microM) of ATRA derepresses transcription by releasing CoR from, and recruiting CoA to PML-RARalpha, whereas As(2)O(3) triggers a rapid degradation of PML-RARalpha. In fact, the two drugs act on the same oncoprotein through targeting different moieties and in distinct ways and thereby abrogate its dominant-negative effects on regulatory pathways necessary for granulocytic differentiation. As to apoptosis, it is clear that high-dose As(2)O(3) can induce mitochondria-mediated cell death pathway in a thiol-dependent manner, while the mechanism of ATRA-induced apoptosis needs further elucidation. Transcriptomic and proteomic analysis are also expected to find new molecular targets. It is the hope that what we have learnt from APL will benefit further developments of anti-leukemia therapy.