We propose research and development to optimize a multimedia approach we have taken to helping children build reading fluency. In this e-book application we highlight text phrase by phrase in synchrony with a fluent voiceover. Children reading in this environment have shown substantial increases in reading fluency. New research coordinating the study of eye movement with detailed study of oral reading fluency, however, shows that shifting the offset between phrasal text highlight and the beginning of voiceover will change the aspect of reading that is being assisted. Minimal offset reinforces word recognition; increased offset reinforces oral reading prosody. Study of fluent readers shows that the actual offset between fixation and phonation changes as a function of several aspects of text. This project will document that changing offset and build a dynamic model of it that can be incorporated into an assisted reading application that helps children increase their oral reading fluency. We anticipate that this project will result in a statistical model and prototype for controlling the synchrony of text highlighting and voiceover audio in a multimedia assisted reading application. Users will be able to control the synchrony thus giving the application a broad market across readers of differing fluency development. The resulting commercial application will have broad appeal and marketability. We have found that even without dynamic control of offset, children become highly engaged in this multimedia e-book environment and that their reading prosody and fluency improves rapidly and markedly. What we are proposing here will improve the user¿s experience and performance.