The Ministry of Transport is planning for the construction of new roads in its territory. Many projects are being considered, and the Ministry needs to identify the worthwhile projects for which the benefits exceed the costs. Among costs and benefits are the expense of constructing the road, the time saved by motorists using the new road rather than some other road, the time saved through the reduction of congestion on other roads, and the expected increase or decrease in the number of deaths due to traffic accidents. I am concerned in this paper with the last item on the list. It is, of course, difficult to estimate fatalities accurately, but this difficulty is neither unique nor central to our problem. The problem is how best to place a monetary value on fatalities, so that the expected number of fatalities can be compared with other costs and benefits in deciding whether to build a road. However chosen, the monetary value of fatalities for use in this context is what the economist means by “the value of life.” The term is almost a joke, a bit of gallows humor to exorcise the ghoulishness that inevitably clings to analysis of life and death in monetary terms.