The thesis reports six experiments investigating word recognition and reading in Arabic. Experiment 1 looked at the word superiority effect in Arabic word recognition using brief presentations of stimuli (five-letter real words, pseudo-words, non-words, and inverted real words) in a Reicher-Wheeler task. The results of this experiment showed advantages for the recognition of words over pseudo-words and illegal non-words, and for pseudo-words over illegal non-words. Experiment 2 was a follow-up experiment that also examined the word superiority effect in Arabic by using the lexical decision task. In this experiment, participants viewed briefly presented real words and legal non-words, with the results showing that Arabic real words were recognised quicker and more accurately than non-words. Experiment 3 investigated the landing position effects for three, five, and seven letter words in Arabic using eye movements while reading. The results showed that the preferred viewing location (PVL) is at the right of centre of words in Arabic, similar to that for Hebrew. Experiment 4 re-examined the optimal viewing position in Arabic word recognition using five-letter Arabic words and non-words in a lexical decision task. The results showed that participants recognised words most quickly and most accurately when fixating inter-letter locations at the middle of words, indicating that the OVP for Arabic word recognition is at a word's centre. Experiment 5 used the Reicher-Wheeler task and Experiment 6 used the lexical decision task to re-examine the claim that an anatomical division in the human fovea has consequences for word recognition. The findings revealed the superiority of the right visual field for words displayed outside the foveal and no asymmetries for words displayed within foveal vision. Thus far the research has made an important advance on our understanding of processes involved in Arabic word recognition by revealing that word superiority and pseudo-word superiority effects similar to those reported in Latinate languages are also observed in Arabic, and that the OVP effect in Arabic differs from that found in English. The reading results indicate that, similar to other languages, parafoveal word length information is used to guide saccade targeting in Arabic.