Fluency is an essential feature of skilled reading. Reading speed is an overt reflection of automaticity in decoding and similarly deft control over other component skills. There is considerable support for oral reading speed as a valid and reliable indicator of general reading skill. In fluent reading, whether oral or silent, processes of word recognition, comprehension, and eye movement control must be tightly coordinated, as characterized in several implemented models of eye movements over print (Reichle et al., 2013; Reilly & Radach, 2006; Engbert et al., 2005). Oral reading, in addition to those requirements shared with silent reading, places an additional demand on the reader: the need to produce accurate, prosodically appropriate speech in conjunction with the reading process itself. Both oral and silent reading rely not only on good control of component processes, but also on facile coordination of those processes (Berninger et al., 2001; Breznitz, 2003). Rapid automatized naming (RAN) is another fluencylike predictor of reading skills (see Bowey, 2005; Kirby et al., 2010; Hulme & Snowling, 2012; Cutting & Scarborough, 2006; Wolf et al., 2000). RAN’s task demands overlap considerably with reading, especially oral reading (Denckla & Rudel, 1974). Yet, the nature of the connections between the two capabilities remains uncertain. This research exploits commonalities between reading and RAN to help resolve uncertainties in their causes. We do so by working toward a unified, theoretical account of performance across tasks. Our main purpose is not to explain RAN as such, but a model of reading fluency that explicitly handles RAN will have distinct advantages over one that does not. It will be constrained by the need to accommodate both tasks, and so may provide insight into relationships between the two that will prove helpful in refining practical uses of RAN as a reading assessment. Our approach is two-fold. First, we will assess relationships among oral and silent reading, and standard indicators of reading skillin three critical periods in the development of skilled reading: a learning-to-read group (7-8yo), a transition-to-fluency group (10-11yo), and a full-literacy group (17-18yo). We will evaluate developmental differences between groups and individual differences within groups. We will apply new ways to investigate RAN performance and oral and silent reading fluency derived from analysis of gaze behavior, and of asynchronies in gaze and speech behaviors produced during these tasks (Jones et al., 2008). We make novel use of speech processing technology to facilitate alignment of speech and gaze signals to accommodate the large volume of speech data generated by the project. Second, we will implement a unified model of coordination among component processes for oral and silent reading(Reichle et al., 2009; Reichle, Pollatsek, et al., 2012). A successful unified model will serve as a testbed for investigating structural relationships among oral and silent reading, RAN and component skills.