Antibody-based therapeutic approaches are yielding more and more of the promise they have held since the conception of the ‘magic bullet’ theory by Paul Ehrlich. The beneficial effect of antibody-based therapies is directly related to antibody-dependent functions, such as neutralization and antibody-dependent cellular cytotoxicity, but in many cases also relies on the delivery of toxic compounds to cancerous cells. However, the clinical utility of toxic antibody conjugates can be significantly hampered by side effects. Ideal effector compounds are inactive ‘en route’, but gain full activity once the antibody conjugate has bound to cancerous cells. Of significant potential in this respect are the pro-apoptotic ligands Tumor Necrosis Factor (TNF), fibroblast-associated cell-surface ligand (FasL) and TNF-related apoptosisinducing ligand (TRAIL). TNF ligands are normally present as homotrimeric transmembrane proteins, but can also be processed into a soluble trimeric form. Compared to their corresponding transmembrane counterpart, soluble TNF, FasL and TRAIL have a strongly reduced capacity to activate TNF receptor 2, Fas and TRAIL receptor 2. However, all sequence information required for full activation of these receptors is latently retained in these soluble ligands and can be unmasked by oligomerization or cell surface immobilization. The latter provides a clear rationale for the use of these ligands as effectors in antibody-based therapy. The antibody-targeted ligand will be in a relatively inactive soluble form while en route. However, once bound to the targeted cancer cell the soluble TNF ligand fusion proteins will be converted into fully active membrane ligand-like molecules. Here we will, after briefly detailing the biology of TNF, TRAIL and FasL, focus on the promises and pitfalls of targeted TNF ligand fusion proteins in achieving a ‘magic bullet’ with maximum cancer selective activity and minimal side effects.