In this paper we present welfare estimates from a contingent valuation (CV) study which investigates the potential benefits derived by tourists from the implementation of a programme aimed at preserving the traditional agricultural landscape in a typical Mediterranean area: the National Park of Cilento (Southern Italy). Here, under current market conditions, farming activities are being gradually abandoned. As a result, the alternative to the cultivated landscape is a much less appealing one, where the various stages of progression of land abandonment dominate. To ease the cognitive task of respondents, CV responses were elicited using the discrete choice referendum format. To supplement the inherent inefficiency of discrete choice responses one follow-up question was also administered. The sample responses are analysed by three methods. First, by a log-normal model which allows a random utility interpretation. Then by a series of beta models, which require the definition of the maximum in the range of willingness to pay and reflect a purely statistical approach. Finally, by means of the non-parametric Kaplan-Meier-Turnbull probability estimates, which is robust to potential parametric mis-specifications. The welfare estimates obtained by various methods were similar and approached one Euro per day-visit. Our results indicate that referendum CV provides plausible estimates of WTP for agricultural landscape conservation from the tourists' population. From a conservative inference it appears that in 1997 the traditional farming produced a landscape externality for tourists which reached at least 8 million Euro. Provision of landscape is only one of many unremunerated activities provided by farmers, so more research should be aimed at valuing public goods produced by farming in recreationally valuable areas and elsewhere. We argue that the policy tools currently employed in the European Common Agricultural Policy, for the purpose of rural landscape preservation, are inadequate in the context of typical Mediterranean agriculture. Here the main cause of rural landscape deterioration seems to be the abandonment of agricultural production, rather than its intensification, so the main features of interest are other than those currently protected by EU policy.