The concept that bacterial infection could cause cancer has only recently become accepted because of the strong epidemiological and molecular evidence for a major carcinogenic role played by Helicobacter pylori. However, information on other potential bacterial carcinogens is very limited and thereby unconvincing. A different approach is to assess bacteria for potentially pro-carcinogenic properties. The Pasteurella multocida toxin (PMT) has many properties that mark it out as a potential carcinogen. PMT is a highly potent mitogen and has been demonstrated to block apoptosis. PMT modifies and activates members of three of the four families of heterotrimeric G-proteins, all of which have potential roles in carcinogenesis. Many signalling components downstream of these G-proteins are known proto-oncogenes and have been shown to be activated by PMT. These include, amongst others, the Rho GTPase, focal adhesion kinase, cyclooxygenase-2, β-catenin signalling and calcium signalling. PMT action potentially influences many of the acquired Hanahan/Weinberg capabilities necessary for oncogenic transformation. Although there is little evidence that PMT might have a role in human cancer, it serves as an important and novel paradigm for a bacterial link to cancer.