The present study is a corpus-based grammatical investigation for the computational identification of metaphors. The aim of the study is to set a grammatical criterion for the computational identification of metaphors in the Holy Qur'an and propose a computer software input rule for the grammatical identification of metaphorical candidates. The work presented in this study draws on the cognitive theory of metaphor (Lakoff and Johnson, 1980), authentic exegeses of the Holy Qur'an, and finally and most significantly it draws on and extends techniques from computational linguistics. The corpus of the study is one surah from the Holy Qur'an, Surat Hud. A methodology for the investigation of candidate metaphor grammatical structure was developed to deduce the grammatical markers of metaphor and utilize them in the computational identification of metaphors. A quantitative and qualitative interpretation of these results and how they can contribute to the computer software suggested for a computer identification of metaphor in the Holy Qur'an is made. The study ended with a theoretical framework that was applied to the corpus to find metaphors by findings specific grammatical markers.