Modality is a semantic medium that colors the way the language user views the world around him/ her in terms of certainty, necessity and obligation; hence, it places extra effort on the translator while attempting to capture modalistic shades of meaning. The task may become more challenging when the translator is dealing with a language pair where modality is grammar-oriented in one member (English, for example) and lexis-oriented in the other (Arabic, for example). The present paper aims to investigate the rendering of speaker participation in the speech event as embodied in modality when translating English fiction into Arabic. In particular, it will examine the corpus of two sets of data involving past modality (modal + have + past participle) extracted from two English novels which will be compared with their counterparts in the Arabic translations. Four main issues will be discussed. The first is to see whether the distinction between epistemic and deontic modality is maintained in translation. The second is to check whether the translators are sensitive to the import of modality in discourse as manifested in the speaker's attitudes toward what is happening. The third is to check whether English modalized propositions are sometimes erroneously rendered into modality-free Arabic propositions. Last, the study discusses the Arabic modality markers employed to capture past modality. Both a quantitative account (focusing on form and function) and a qualitative analysis (focusing on adequacy of translation procedures) are furnished.