In this study, students in Grade 2 read different sets of words under 4 experimental training conditions (context/feedback, isolation/feedback, context/no-feedback, isolation/no-feedback). Training took place over 10 trials, followed by a spelling test and a delayed reading posttest. Reading in context boosted reading accuracy initially; in contrast, the external support garnered from feedback resulted in heightened reading accuracy throughout training, as well as 1 week later. Different patterns were noted in spelling transfer; first, there was no effect of reading feedback on spelling at posttest. Second, the highest spelling scores were observed when children practiced reading words in isolation versus in context. In sum, providing feedback and/or context helps children read words accurately, which in turn seems to create orthographic representations that are "good enough" to support reading accuracy. However, reading in isolation seems to produce orthographic representations that are higher in quality, and therefore better able to support precise spelling.