Impact assessment studies consistently show that the benefits generated by plant breeding are large, positive and widely distributed. Numerous case studies have concluded that investment in plant breeding research generates attractive rates of return compared to alternative investment opportunities, that welfare gains resulting from the adoption of modern varieties (MVs) reach both favoured and marginal environments, and that benefits are broadly shared by producers and consumers. But just how reliable are the results of studies that estimate the benefits of plant breeding research? This article reviews methods used to estimate the benefits of plant breeding research and discusses theoretical and empirical issues that often receive inadequate attention in applied impact assessment work. Our objective is not to question the validity of the theoretical frameworks commonly used to estimate the benefits of plant breeding research, but rather to examine problems that can arise when the widely accepted theoretical frameworks are used for empirical analysis. Most of these problems can be grouped into three basic categories: ( 1) problems associated with measuring adoption and diffusion of MV s, (2) problems associated with estimating benefits attributable to adoption ofMVs, and (3) problems associated with assigning credit among the various plant breeding programmes that participated in developing the MVs. © 2003 Elsevier B.V. All rights reserved.