Evaluation of recently introduced policy instruments (RIPIs) is especially problematic, because only some effects have occurred, and information on them is imperfect. Policy makers and the public at large are, however, particularly interested in early evaluations. This article examines problems with the retrospective evaluation of RIPIs, and explores the advantages of using intervention theories in these evaluations. Two case studies from the field of environmental policy instruments are used as examples. It is argued that when evidence on final outcomes is largely unavailable, an intervention theory is a useful tool to overcome information problems. By using intervention theories, it is possible to identify observable prerequisites that precede intended, but not yet occurred, outcomes.