This paper examines the value to Swedish citizens of reducing the risk for salmonella bacteria in chicken filet. The contingent valuation (CV) study is based on the results of a postal questionnaire that was distributed to 2 000 randomly selected Swedish citizens aged 18-74. The valuation format used is a stated preference double bounded dichotomous choice. We employ the non-parametric Turnbull Lower Bound method in combination with Monte Carlo simulations to obtain lower bound estimates of the mean and median values of expected willingness-to-pay (WTP) for reducing the risk for salmonellosis, as well as values of a statistical case (VSC) and a statistical life (VSL). We find a VSC of between SEK 121 045 (110 297–131 814) and SEK 182 966 (167 915–197 896) depending on the format used (values in parentheses constitute a 90 percent confidence interval). VSL values of SEK 13.3 million and 48.3 million are estimated using different formats, but neither estimation is statistically significant. Since this is the first Swedish study on WTP for food safety, mean and median values of VSL and VSC cannot be directly compared with previous results, but the values obtained are in line with comparable Swedish studies on WTP for traffic safety as well as with international studies related to food safety. We do not find any strong linkage between WTP and income, age or gender. Scale sensitivity seems to depend on which model is chosen, while household size, risk perception ability and perceived Quality Adjusted Life Years (QALY:s) lost seem to be strong predictors of WTP.