Over the last decade, the ability to simulate the results of NDE has made major strides. A number of tools are now available, in various stages of development, that can accurately predict the results of rather practical measurements made on parts with complex geometries and material properties. This paper documents the results of a discussion session in which ways to best take advantage of this technical capability were considered. Included are considerations of how to most effectively couple to users, points in the life cycle of components when simulations can be most effectively used, the role of models in assessing trade‐offs between various strategies for assuring reliability, and the role of models in documenting the economic benefits of NDE. Also considered are certain strategic issues concerning model development. © 2004 American Institute of Physics