The main aim of this study is to develop a spell-checker system for Arabic language. This is done by investigating the viability of applying the radix search tree approach. Through this scientific research several shrubs that represent Arabic characters will be built through serialized tracking of characters word where it can be added to the dictionary and with a special mark in the node that contains the last characters from each word; on other side during searching process, every word can be tracked character by character according suitable path inside its shrub, Accordingly, correct word can be recognized if and only if searching process locates some leaves during the traverse of the shrub. Otherwise, the word will be considered incorrect.