Reading comprehension, defined as the active extraction and construction of meaning from all kinds of text, requires children to fluently decode and understand what they are reading. Basic processes underlying reading comprehension are complex and call on the oral language system and a conscious understanding of this system, i.e., metalinguistic awareness, at all levels from semantic and morpho-syntactic to pragmatic awareness. Higher order meta-cognitive skills also appear to contribute to comprehension. Thus there is emerging research for what is being called the usual suspects. These include semantic knowledge and vocabulary, comprehension strategy use, awareness of text structure, background knowledge, and self-regulation, including attention among others. Even with the usual suspects, however, there is much to be learned. For example, the association between use of non-mainstream American English (NMAE), such as African American English, and literacy has been documented in correlational research but it is not clear that there is a causal association. Indeed, recent research suggests that it is not NMAE use per se that is related to reading but rather the use of MAE rather than NMAE in contexts where it is expected, such as school. New findings suggest that those students who continue to use NMAE in situations that expect the use of School English (i.e., MAE), such as writing tasks, past second grade are at serious risk for reading underachievement. This design study was designed to test whether second through fourth graders could be taught to use School English (i.e., MAE) rather than Home English (NMAE) when writing (a context where School English is expected), and whether there was an advantage to providing an explicit focus on the contrasts between Home English and School English. The design experiment was conducted at schools in North Florida where at least 50% of the students qualified for the US Free and Reduced Lunch program, a frequently used marker of family poverty. The schools were also racially and ethnically diverse. Second through fourth grade students (n = 116) who used at least one feature of NMAE on the Diagnostic Evaluation of Language Variation Screener (DELV-S) part one, an oral language assessment, or in a written narrative participated in the study. For this design experiment, the authors developed a short 4-week intervention. Two important findings emerged from this study: First dialect shifting is malleable. Overall, students in both treatment conditions demonstrated greater knowledge and use of three MAE/NMAE contrastive features in written tasks than students in the control group. Second, explicitly contrasting home and school language enhanced dialect-shifting instruction. Students in the Dialect Awareness treatment group had significantly better outcomes on all post-intervention measures than either the editing-only treatment group or the control. Figures are appended.