. The aim of the present paper is to compare productivity of some locative suffixes in Persian using a corpus-based approach. answering the following questions is focused: a) what is the hierarchy of productivity of target locative suffixes? b) How does productivity of target suffixes correlate to semantic predictability of derived words? In total, there are 347808 tokens and 30 types derived from suffixes /-lɑχ/, /-setɑn/, /-gɑh/, /-zɑr/, /-dɑn/, /-særɑ/ and /-kæde/ in Hamshahri corpus. Results suggest that /-dɑn/ and /-lɑχ/, in order, are the most and the least productive suffixes. productivity of target suffixes is in this order:

/-dɑn/> /-zɑr/> /-kæde/> /-gɑh/> /-særɑ/> /-setɑn/> /-lɑχ/

It should be mentioned that difference between productivity of /-gɑh/ an /-særɑ/ is not significant. Also, it was shown that there is no correlation between degree of productivity of target suffixes and semantic predictability of derived words.