Hillinger (198) reported that a target word was identified as a real word faster when it was preceded by a phonetically related prime than by a neutral prime. He found that this facilitation occurred even when the prime and target were graphemically dissimilar. On the basis of this he concluded that activation spread automatically from the lexical entry of the prime to phonologically related lexical entries. The series of experiments reported in this thesis were designed to investigate several aspects of Hillinger's experiments. Hillinger's findings were replicated when his design was employed. Further investigations showed that minor variations in this design resulted in very different patterns of facilitation and inhibition. While it was demonstrated that lexical decisions to word targets are facilitated, apparently automatically, when the targets are preceded by graphemically similar-phonemically similar primes, no such facilitation was observed when graphemically dissimilar-graphemically similar primes were used. These results were found even when subjects were required to access the phonology of the primes. It was concluded that phonetic facilitation is the result of a process of structure that is based both on phonology and orthography.