Markets for non-timber tree products (NTTPs) are widely viewed with great promise because of their potential for achieving a range of development objectives. Markets can, of course, be influenced by policy, but in order to do so, market information is needed to better inform the policy process. There has been increasing concern that few market studies have been adequate for doing so. There has been ambivalence at the policy level because these markets are highly diverse and difficult to characterize, and because governments tend to view NTTP markets as a threat to the conservation and management of forests and woodlands. Perhaps most importantly, however, there has been a failure to link the design of market studies with the potential for policy change. We argue in this paper that policies per se can be used to influence the operation of these markets, and that good market research should be more carefully designed to reflect explicit policy objectives.