Despite a large body of evidence investigating how letter-position information is encoded during lexical-processing during both isolated-word and reading paradigms, it is still not clear whether the mechanism to encode letter-position information is modulated by age and/or reading ability. The aim of the present research was to investigate developmental changes in letter-position encoding during reading. The first experiment investigated the influence of letter-position encoding on the time course of lexical and post-lexical processing during reading. It examined whether the prior exposure of a word’s transposed-letter neighbour (TLN) earlier in the same sentence interfered with that word identification in skilled-adult-readers. Results showed that in skilled-readers, TLNs caused a target words’ misidentification, triggering post-lexical strategies of checking. The second experiment investigated whether children extracted letter-position information independently from letter-identity information from the parafovea as adults do. Results showed that although children had longer reading times overall than adults, both adults and children pre-processed orthographic information from the parafovea and encoded letter-position information using a spatial coding mechanism. Finally, the third experiment examined whether children’s reading ability influenced letter-position encoding during lexical-processing in reading. Adults, skilled and less-skilled child readers read sentences with words containing two letters-transposed (positions 1&2, 1&3 and 2&3). Results showed that words with transposed-letters in position 13 caused the most disruption to reading, while words with transposed-letters in position 23 caused the least disruption to reading in both adults and skilled child readers. Less skilled child readers showed that although they showed disruption when identifying transposed-letters nonwords, the cost did not vary systematically depending on the letter-position. This suggests that less skilled child readers with fewer, high-quality lexical representations were activating phonological codes for word identification via the fine-grained route. In contrast, both adults and skilled child readers with more, high quality lexical representations were activating orthographic codes for word identification via the coarse-grained route.