River restoration has received nationwide interest and financial investment, yet thousands of restorations take place annually, only a fraction of which are evaluated and benefit from the integrated insights of practitioners and scientists.  The University of California joined the National River Restoration Science Synthesis to gather and analyze information about current design, implementation, and monitoring practices used in California to inform and improve future practice.  In the first two phases of this study, we compiled a database of ~4,000 restoration projects undertaken through 2003, and conducted follow-up interviews with implementers of randomly-selected projects. In the third phase, we conducted 40 systematic post-project appraisals of restoration projects to compare current and pre-project conditions (where available), and summarize performance indicators.  Surveys in the field (including cross-section and long-profiles, pool-riffle sequences, vegetative cover and composition, bentho-macroinvertebrate surveys, and stakeholder interviews) enabled us to document landscape change and response. We found that many California stream projects have conflicting goals; that interventions are moving away from traditional “hard” engineering approaches; that stakeholder participation is a crucial link to restoration effectiveness; and that monitoring is far more widespread than predicted, but objective and long-term assessment of completed projects hinders the advance of restoration science.