There is a dearth of scientific literature on speech recognition for Arabic language. Moreover, most of the research in this area is focused on using Romanized corpus which results in poor accuracy of speech recognition systems. This book presents a study of speech recognition for the Arabic language using a fully-diacritized corpus of an old classic Arabic book called Sahih Al-Bukhari. After comparing the accuracy of the Arabic speech recognition systems using the two types of corpora, this study has shown that better accuracy is achievable using fully-diacritized corpus rather than Romanized corpus. This conclusion can be used as a first step towards developing more accurate Arabic speech recognition systems and to encourage researchers to build extensive library of fully-diacritized corpora of old Arabic books.