RealTime Physics (RTP) and Interactive Lecture Demonstrations (ILD) have had great success in improving student learning at their development sites, but dissemination of research-based curriculum materials often shows that student learning at secondary implementation sites is less than at primary, development sites. Five schools using RTP and ILD materials were investigated during the first three years of the implementation of the materials. Student learning of the relevant physics was investigated through the use of several standardized tests (including the Maryland Physics Expectations Survey, the Force and Motion Concept Evaluation, and the Electric Circuits Concept Evaluation). Results show that normalized gains of pre- and post-test scores are larger for classes effectively implementing the RTP and ILD materials than for classes using the regular instruction methods or having serious implementation problems. Furthermore, post-test scores are not as high as reported by the development sites.