Almost ever since their invention, monoclonal antibodies have held the promise of cancer-specific drug targeting--Paul Ehrlich's "magic bullet"--but only during the past decade have a modest number of anti-cancer antibodies received approval for clinical use. These, however, have proven largely successful, with very different kinds of conventional or recombinant, murine, humanized, recombinant fully human and fusion constructs, and mechanisms of action as diverse as complement or antibody dependent cytotoxicity, anti-angiogenesis, and growth factor inhibition. In these latter two mechanisms of action, antibodies compete with novel small-molecule drugs. This review tries to elucidate current trends in those antibody-based therapeutics that are currently in clinical development. With more than 400 such molecules registered for clinical trials, it is far from a chance to be complete. Still, from those antibodies selected for a closer view, two large trends can be distilled: The movement towards increasingly molecularly defined recombinant constructs, and away from classical antibody effector functions in immune activation towards additional mechanisms of action--either by stimulation or (more often) inhibition of a molecular target function, or by additional functional moieties attached to the antibody scaffold. While these trends probably mark the future of antibody development for cancer therapy and clinical applications in general, a considerable number of more conventional--hybridoma generated or recombinantly chimerized or humanized--Fc receptor-activating antibodies, originally generated a decade or longer ago, continue to make their way through clinical trials, some with remarkable success.