Anthrax toxin, the major virulence factor of Bacillus anthracis, consists of three polypeptides: protective antigen (PrAg), lethal factor (LF) and oedema factor (EF). To intoxicate mammalian cells, PrAg binds to its cellular receptors and is subsequently activated via proteolysis, yielding a carboxyl-terminal fragment which coordinately assembles to form heptamers that bind and translocate LF and EF into the cytosol to exert their cytotoxic effects. Substantial progress has been made in recent years towards the characterisation of the structure and function of anthrax toxin, and this has greatly facilitated rational drug design of antianthrax agents. There is also emerging evidence that toxins can be manipulated for cancer therapy. LF can efficiently inactivate several mitogen-activated protein kinase kinases (MAPKKs) via cleavage of their amino-terminal sequences. Consequently, antitumour effects of wild type lethal toxin were observed after treatment of mitogen-activated protein kinase (MAPK)-dependent tumours such as human melanomas. Modification of the toxin's proteolytic activation site limits its cytotoxicity to certain cell types and creates a versatile method of treatment. One approach that has successfully achieved specific tumour targeting is the alteration of the furin cleavage of PrAg so that it is not activated by furin, but, alternatively, by proteases that are highly expressed by tumour tissues, including matrix metalloproteases and urokinase.