This study analyzes stated willingness to pay (WTP) for traffic safety, the use of traffic safety equipments, and the consistency between the two. Using data from a Swedish contingent valuation study we find that the estimated value of a statistical life (VSL) based on the respondents' rear-seatbelt usage is similar to the estimate found using the respondents' stated WTP. However, when estimating VSL based on the respondents' use of bicycle helmets we find a significantly higher VSL; the VSL from bicycle-helmet usage is 7 times higher than the estimate based on seatbelt usage. Moreover, we do not find any strong relationship between risk perception and usage, or individual stated WTP and usage. Hence, the main conclusion, based on our analysis, is that stated and observed WTP are not consistent.