The current study seeks to clarify various ways in which metonymy can affect the meaning of Persian noun-noun nominal compounds. An analysis of 280 endo- and exocentric Persian noun-noun nominal compounds reveals that as far as the role of metonymy in the construction of meaning is concerned, this cognitive mechanism can affect the meaning of Persian nominal compounds in four ways as follows: a) metonymical modifier, b) metonymical head, c) metonymical head and modifier, and d) metonymic nominal compounds as a whole. While the pattern of metonymical modifier only affects the meaning of endocentric compounds, the other three patterns function in the meaning of exocentric compounds. This study substantiates Brdar and Brdar-Szabo’s (2013) and Brdar’s (2017) assertion that metonymy may either act upon the constituents of the compound, i.e., before compounding (the first, second, and third pattern), or the compound as a whole, i.e., after the combination of constituents (the fourth pattern). It is also argued that the metonymical relationship between head and modifier cannot be regarded as a pattern. If such an argument was plausible, non-figurative endocentric compounds would also be metonymical and compounding would basically require the functioning of metonymy.The current study seeks to clarify various ways in which metonymy can affect the meaning of Persian noun-noun nominal compounds. An analysis of 280 endo- and exocentric Persian noun-noun nominal compounds reveals that as far as the role of metonymy in the construction of meaning is concerned, this cognitive mechanism can affect the meaning of Persian nominal compounds in four ways as follows: a) metonymical modifier, b) metonymical head, c) metonymical head and modifier, and d) metonymic nominal compounds as a whole. While the pattern of metonymical modifier only affects the meaning of endocentric compounds, the other three patterns function in the meaning of exocentric compounds. This study substantiates Brdar and Brdar-Szabo’s (2013) and Brdar’s (2017) assertion that metonymy may either act upon the constituents of the compound, i.e., before compounding (the first, second, and third pattern), or the compound as a whole, i.e., after the combination of constituents (the fourth pattern). It is also argued that the metonymical relationship between head and modifier cannot be regarded as a pattern. If such an argument was plausible, non-figurative endocentric compounds would also be metonymical and compounding would basically require the functioning of metonymy.