Fallout distributions are calculated for nuclear attacks on the contig
uous United States. Four attack scenarios are treated, including count
erforce and counterforce-countervalue attacks, for meteorological cond
itions associated with a typical day in summer and one in winter. The
countervalue attacks contain mostly airbursts. To determine fallout ef
fects, the population surviving the prompt effects is first calculated
. For the prompt effects, a ''conflagration'' type model is used. The
counterforce attack produces about 8 million prompt deaths, and the co
unterforce-countervalue case projects 98 million prompt deaths. Partia
l relocation before attack to low-risk fallout areas at least 15 km fr
om potential strategic targets would result in a decrease in projectio
ns of deaths by tens of millions. For fallout risk calculations, only
the dose received in the first 48 hours (the early, of local fallout)
is considered. Populations are assumed to be sheltered, with a shelter
protection factor profile that varies for a large urban area, a small
urban area, or a rural area. With these profiles, without relocation,
the fallout fatalities for all four attack scenarios are calculated t
o be less than one million people. This can be compared to fallout fat
alitites of about 10 million for a hypothetical unsheltered ''phantom'
' population.