The accidents at the Three Mile Island (TMI) and Chernobyl nuclear reactors have triggered particularly intense concern about radiation hazards. The TMI accident, in Pennsylvania in 1979, resulted from a combination of deficient equipment and operator errors. Even though there was severe damage to the nuclear fuel within the reactor, very little radioactivity escaped into the outside environment. The effectiveness of the large concrete containment building that surrounded the reactor contributed to this relatively small release. Subsequent studies concluded that the maximum dose received by any member of the public was under 1 mSv (100 mrem). The collective off-site dose is estimated to have been about 20 person-Sv. Under the standard low-dose assumption, this corresponds to one eventual cancer fatality in the neighboring population of 2 million people. (This population receives an annual collective dose of about 6000 person-Sv from natural sources.)

The 1986 Chernobyl accident was far more serious. It occurred in a reactor with an unsafe reactor design unique to the Soviet Union. The reactor had no effective containment, and there was a very large release of radionuclides to the environment. The accident led to the death within several months of 31 reactor personnel and firefighters–28 from a combination of radiation effects and burns from fire, 2 from other injuries, and one from a heart attack. A total of 237 workers was hospitalized for symptoms of radiation sickness, including the 28 who died. A 1996 summary reported an additional 14 deaths among the more severely exposed workers, but it is not clear that these deaths were all due to the prior exposure.

There is strong evidence of a substantial increase in thyroid cancers among children living in the vicinity. No other health effects from Chernobyl have been convincing established. However, it is too soon for them to have been fully manifested. Standard calculations of radiation effects predict that there will be a large number of excess cancer deaths among the so-called "liquidators," who were engaged in cleanup operations after the accident, as well as in the neighboring population.

Considering impacts at greater distances, one early study estimated that the collective dose in the Northern Hemisphere over a 50-year period would be about 930,000 person-Sv. While there is substantial uncertainty in the dose estimate, there is even greater uncertainty as to the impact. If one accepts the linearity hypothesis and assumes 0.05 fatalities per Sv, this dose corresponds to 47,000 eventual cancer fatalities. About 29,000 of these fatalities would occur in Europe (outside the former Soviet Union) due to a cumulative collective dose of 580,000 Sv–an average individual lifetime dose of 1.2 mSv for 490 million people. Given these low average doses, any estimate of predicted deaths from Chernobyl is highly speculative. The deaths will not be identifiable, being masked by the 88,000,000 "normal" cancer fatalities expected in this region during the 50-year period.