The use of survey or contingent valuation methods to estimate an individual's valuation of non-marketed goods, especially environmental resources, has attracted increasing attention in recent years. Initially, research efforts in this area were viewed by a majority of the economics profession with considerable skepticism. However, with the increased need for information on individuals’ valuation of a whole range of environmental resources and limitations on the ability of indirect market-based methods for valuing all of these resources, there has been a substantial increase in the use of contingent valuation methods to provide this information. Indeed, a recent state-of-the-art assessment (see Cummings et al. [1984]) of the contingent valuation method (CVM) has been able to develop a set of reference operating conditions under which it was reasonable to expect the CVM approach would yield estimates with accuracy that was approximately comparable to the indirect methods. Clearly the definition of these conditions is a judgmental one. Nonetheless, it was based on a substantial number of comparative studies evaluating the relationship between CVM and indirect market estimates of the benefits associated with changes in specific environmental resources. Moreover, it does reflect the changing attitude toward the CVM approach. It is therefore particularly appropriate to consider new directions in the development of the contingent valuation method.