Animal welfare presents particular policy challenges. Good welfare provides private productivity benefits to producers and some level of positive external benefits to people who care about animal welfare status. In enacting welfare legislation and setting regulatory standards, government needs to measure costs and benefits of welfare changes. While costs are generally observable, the nature of market failure means that welfare benefits are not truly observed in welfare related transactions. Accordingly non-market benefits assessment methods are required to measure the total economic value of welfare improvement. This paper compares the results of two stated preference methods to measure the policy benefits of the proposed EU broiler Welfare Directive. Contingent valuation presents the welfare improvement as a policy bundle and elicits willingness to pay in a referendum or one-off purchase decision. Choice experiments break down the welfare good into its constituent attributes, which may be of interest in designing policy. The methods provide divergence aggregate benefit estimates, which are an artefact of the methodology and the payment methods.