A recent U.S. Office of Education-sponsored study of the overall impact of educational innovation, the Longitudinal Study of Educational Practices (Project LONGSTEP), showed that student achievement was not related to emphasis on innovation in any substantial or consistently positive way. These findings have been mistakenly interpreted by some as a sweeping criticism of all innovative school practices. A more accurate interpretation, however, is that new educational techniques per se will not automatically guarantee dramatic improvements in achievement. This paper reviews Project LONGSTEP's results and discusses the major policy implications of these findings for decision-makers in both education and government. (Author)