PURPOSE: To compare the potential of five functional magnetic resonance imaging (fMRI) language paradigms in activating language areas in Persian-speaking volunteers in order to optimize these tasks for clinically useful protocol. MATERIALS AND METHODS: 16 healthy right-handed Persian-speaking volunteers were studied. Each individual performed five tasks during the fMRI scan: word generation (WG), object naming (ON), word reading (WR), word production (WP), and reverse word reading (RWR). The ability of each task to activate classical language areas was assessed using group analysis. In addition, the lateralization index (LI) for each subject-task was calculated and compared. RESULTS: We found that WP, RWR, and WG robustly activated language-related areas in the dominant hemisphere. ON and WR failed to sufficiently delineate these activation areas. Highest activation intensities in the frontal lobe (including Broca's area) were seen with WP, whereas RWR showed the highest LI among all examined tasks. CONCLUSION: Our results demonstrated that the Persian version of WG and newly presented WP and RWR tasks can be reliably used for localization of language-related areas in Persian speakers. This type of language evaluation may be used in presurgical planning of neurosurgical procedures in the Persian population. We recommend application of WP and RWR in future researches establishing the optimized protocol for other language speakers.