Malignant cell transformation is caused by mutations in distinct key regulatory genes involved in cell growth, apoptosis, senescence and differentiation. Particularly in human leukemia, chromosomal translocations involving crucial hematopoietic transcription factors are frequently causally linked to the disease. Transcription factors commonly have a modular structure, comprising distinct domains for DNA- binding, dimerization and protein-protein interaction. Each domain is functionally important and in principle accessible for a molecular-based therapeutic intervention. Uncovering the molecular structure of critical domains will allow the rational development of therapeutic agents that inhibit particular functions of leukemogenic transcription factors. However, so far most approaches are in the experimental stage. Among others, the RUNX1/ETO fusion protein, commonly found within acute myeloid leukemia cells carrying the translocation t(8;21), is currently intensively studied at the functional and structural level as well as in animal models. This combined effort has allowed the development of specific targeting approaches addressing different functional domains of the fusion protein. With a special focus on RUNX1/ETO we will discuss recent strategies to directly interfere with aberrant transcription factors to block their leukemogenic function.