Although most models of reading development present orthographic knowledge as a more advanced and later developing form of knowledge than phonological knowledge, this dissertation presents a model of the development of orthographic knowledge in which generalized orthographic knowledge, the knowledge of symbol patterns within and across words, develops early, at the same time as phonological knowledge and before lexicalized representations of a whole word. However, because phonological and generalized orthographic knowledge are not fully integrated, phonological knowledge masks orthographic knowledge in typical measures of literacy. In study 1 pre-readers' knowledge of the elements that make up words was tested using eye-tracking as a measure of implicit knowledge. We find that pre-reading children as young as 3 have implicit orthographic knowledge regarding the elements that make up words. This supports the prediction that generalized orthographic knowledge develops before lexicalized knowledge. In study 2, children's creative spellings were used to gauge children's implicit knowledge of letter patterns in a naturalistic setting. We find that kindergarteners in particular tend to rely on phonology over orthography when the two are in conflict. This supports the hypothesis that phonological knowledge can mask orthographic knowledge. In study 3, children were asked to decode non-words and their implicit knowledge of letter patterns was measured using eye tracking. I found that early readers show some implicit knowledge when decoding, This supports the hypothesis that generalized orthographic knowledge can be measured in literacy tasks under certain testing conditions. In study 4, children's phonological and orthographic knowledge was tested directly by asking children to sound out and select the best word. Results show that sensitivity to orthographic violations is decreased when phonology is introduced. This is a direct test of the hypothesis that phonological knowledge can mask orthographic knowledge, and findings support this hypothesis. These results suggest that pre-readers show generalized orthographic knowledge before lexicalized knowledge and concurrently with phonological knowledge. Furthermore, this generalized orthographic knowledge initially presents itself implicitly, and in many early literacy tasks the orthographic domain is dominated by phonological concerns. Essentially, orthographic and phonological knowledge develop at the same time; however, until children learn to integrate the two dimensions of written language, they rely on one source over the other.