The major causes of cancer fall into two categories: those over which we have some control and the rest. The latter includes radiation from the Earth that has been a background to human evolution and about which we can do nothing, although we could take more effective steps to limit the accumulation of radon in houses in regions where there is high, localised emission of the gas. The factors that we can control also fall into two groups: those exerting major effects for which the epidemiological evidence is overwhelming and those for which the data are inconclusive and therefore controversial. Of the former, the most familiar is the use of tobacco, which is responsible for 90% of lung tumours. By contrast, there are a number of prominent agents that may have weak tumour-promoting effects. For these the epidemiology is generally unpersuasive and direct experimental evidence has not been forthcoming, as exemplified by the continuing debate over the risks associated with the use of mobile phones. The two most effective measures we could take that would reduce the global cancer burden by at least one third would be to abolish the use of tobacco and to limit the consumption of red meat.