Cancer is one of the longest-known human diseases, yet only in recent times have we begun to perceive that the percentage of neoplasms caused by environmental factors, lifestyle and chemicals, is likely underestimated. The first medical reports associating cancer with pollutants like tars appeared by the early 20th century, but despite initial evidence relating oncogenesis and chromosomal alterations, only after the structure of DNA had been elucidated in the 1950s have genetic disorders been fully perceived as cause. This led to a growing interest in genotoxic and mutagenic pollutants. Even though we are now familiar with a range of environmental carcinogens spanning between aromatic hydrocarbons and asbestos to radionuclides and forms of carbon nanomaterials, establishing causal networks between pollutants and cancer remains cumbersome. In most part, this is due to the complexity of toxicant matrices, unknown modes-of-action of chemicals or their mixtures, the widening array of novel pollutants plus difficulties in subtracting background effects from true aetiology of disease. Recent advances in analytical chemistry, high-throughput toxicology, next-generation sequencing, computational biology and databases that allocate whole normal and cancer genomes, all indicate that we are on the verge of a new age of research into mechanistic ‘oncotoxicology’, but how can it impact risk assessment and prevention?