The regulation of environmental carcinogens has a history of some 50 years as a formal process linking exposure and response. In the past 20 years or so, the regulatory process for these environmental carcinogens has become more closely linked to a risk assessment framework. Although the specific approaches vary across countries and regulatory bodies, there is quite a similarity in overall intent, namely to identify an exposure level that is protective of the public. This cancer risk assessment practice has undergone a significant change in the past 5–10 years whereby a much greater emphasis has been placed on the use of mechanistic data to address uncertainty in risk assessment that includes a move away from default factors. The approach described in the article is based on that developed by the US Environmental Protection Agency for their most recent guidelines, although it is similar to the approaches used by a number of major regulatory agencies. The basis for cancer risk assessments for a specific carcinogen is now a framework that includes a description of the mode of action (MOA) in terms of key events that are required to convert a normal cell to a metastatic tumor. Such an MOA is usually developed for rodents because of a paucity of human tumor data, and so the relevance of this MOA to humans is established using a human relevance framework. Dose–response assessment for tumors (or for key events in the tumor process) uses an extrapolation from data within the observable range to those at environmental exposure levels. The nature of the extrapolation is both complex and somewhat controversial as regards the particular models and approaches that are selected. These range from biologically based dose–response (BBDR) models for situations where mechanistic data are quite extensive to empirical modeling in cases where insufficient mechanistic data are available to support a BBDR approach. The implementation of risk assessment in a regulation is performed using risk management guidelines. These two processes are, perhaps appropriately, being more tightly linked. The future in terms of addressing uncertainty lies in enhancing experimental approaches for better understanding of toxicity pathways and key events in the formation of tumors. This is a need that is being addressed aggressively in many laboratories worldwide.