To study the effect of military service during the Vietnam era on subsequent mortality, we analyzed a randomized natural experiment, the military draft lottery of 1970 to 1972. Between 1974 and 1983, there were 14,145 deaths among California and Pennsylvania men whose dates of birth were in the years for which the draft lottery was held. The group of men with birth dates that made them eligible for the draft had a higher mortality rate than the group with birth dates that exempted them from the draft: suicide was increased by 13 percent (P = 0.005 by two-tailed test), death from motor-vehicle accidents by 8 percent (P = 0.03), and total mortality by 4 percent (P = 0.03). Only 26 percent of the men who were eligible for the draft actually entered the military. If military service (rather than draft eligibility) was the actual risk factor, suicide and death from motor-vehicle accidents would have to have been increased by 86 percent and 53 percent among men who served in the military, to produce the increased risk that we observed among all draft-eligible men. A separate analysis that compared the causes of death in veterans and nonveterans yielded similar estimates: veterans were 65 percent and 49 percent more likely to die from suicide and motor-vehicle accidents, respectively. We conclude that the most likely explanation for these findings is that military service during the Vietnam War caused an increase in subsequent deaths from suicide and motor-vehicle accidents.