In this paper we show that omitted variables and publication bias lead to severely biased estimates of the value of a statistical life. Although our empirical results are obtained in the context of a study of choices about road safety, we suspect that the same issues plague the estimation of monetary trade-offs regarding safety in other contexts.