Cross Language Information Retrieval (CLIR), are aims to retrieve information in one language using queries in another language. CLIR requires that either the query or documents be translated from one language to another. In this work, the basic translation approaches of CLIR were experimented, to select the most appropriate technique in terms of ease of use, efficiency and the volume of available sources needed. Also, different methods for CLIR between Arabic and Hebrew have been investigated. An automated system is devised to accept Hebrew queries, translates the queries, and searches an Arabic document collection to retrieve the most relevant. The system uses dictionary-based approach with English as intermediate language for translating the queries from Hebrew to Arabic. The system was applied in TREC 2002 CLIR standard Arabic Collection Using 50 Hebrew queries (manually translated of Arabic TREC queries). The dictionary used is a constructed bilingual term list includes inflected words, and the translation selection method Top n was applied with synonym operator #wsyn to decrease the ambiguities.