Economic and population growth have led to an unprecedented increase in the value at risk in coastal zones over the last century. To avoid excessive future losses, particularly in the light of projected climate change impacts, coastal zone managers have various instruments at their disposal. These primarily concern land-use planning (establishing buffer zones) and engineering solutions (beach nourishment and coastal protection). In this paper, we focus on risk mitigation through the implementation of buffer zones (setback lines). Foregoing land-use opportunities in coastal regions and protecting coasts is costly, but so is damage caused by inundation and storm erosion. Defining appropriate setback lines for land-use planning purposes is a balancing act. It is, however, unclear what level of protection is facilitated by current approaches for defining setback lines, and whether this is, at least from an economic perspective, sufficient. In this paper, we present an economic model to determine which setback lines would be optimal from an economic perspective. The outcomes of the model provide a useful reference point in the political debate about the acceptability of risk in coastal zones. The main conclusions are: (i) that it is useful to define setback lines on the basis of their exceedance probabilities; (ii) that the exceedance probability of an economically efficient setback line will typically be in the order of magnitude of 1/100 per year; (iii) that it is important to distinguish between situations in which morphological conditions are stationary and non-stationary; and (iv) that long-term uncertainties (eg. due to climate change) influence the exceedance probability of efficient setback lines but only to a limited extent. The economic model stresses the need for a probabilistic approach to beach erosion modelling. The recently-developed Probabilistic Coastal Setback Line was applied at Narrabeen beach, Sydney, Australia, to illustrate how economically optimal setback lines can be derived for specific sites.