Low-levels of dioxin cause cancer in animals. In 1997 dioxin was found to be a human carcinogen by the International Agency for Research on Cancer, based largely on four studies of industrial workers exposed to high levels. Recently there has been interest in estimating human cancer risk at low level environmental exposures. Here we review quantitative exposure-response analyses and risk assessment for low environmental levels based on the largest existing cohort of workers exposed to dioxin (the U.S. NIOSH cohort). We estimate that doubling background levels of exposure, which may occur for example by eating a lot of fish which have accumulated dioxin, will increase lifetime risk of cancer death by 0.1 to 1.0%. In the US the background risk of cancer death by age 75 is 12%, so doubling background levels of dioxin exposure would increase this lifetime risk to somewhere between 12.1 and 13.0%. Our results agree broadly with results from a German cohort, which is the only other cohort for which a quantitative risk assessment has been conducted.