A study of the completeness of death ascertainment through the Social Security Administration (SSA) was undertaken. Groups of subjects of known vital status were submitted by the Los Alamos National Laboratory to the SSA for a mortality record search. Correctly ascertained deaths ranged from 77% to 90% among males and from 53% to 76% among females. A duplicate record search one year later raised correct ascertainment to 88% to 94% of male deaths and to 73% to 87% of female deaths. Death ascertainment appeared to be independent of ethnicity (Anglo and Hispanic), cause of death, location of death, and year of death. Age at time of death was strongly related to death ascertainment. Individuals 65 years of age and over were identified much more accurately than persons less than 65 years of age. Lack of incentive to report deaths of persons under retirement age is believed to be responsible. The age of an industrial facility and of the work force may have a major influence on the adequacy of mortality ascertainment through social security.