To investigate the iron cycle at Station A4 in the Oyashio region of the western subarctic Pacific, we developed a 1-D ecosystem model consisting of 14 components including the iron cycle. The parameters associated with the iron cycle were optimized by assimilating monthly averaged data from time series observations for depth-integrated net primary production, nitrate, silicate, dissolved and particulate iron within the surface mixed layer (ML) and at two depths (200 and 300 m depth). The model successfully reproduced the observations and demonstrated that (1) on an annual basis, winter mixing of subsurface water supplies more dissolved iron (Fed) to the ML than does dust dissolution, (2) Fed concentration in the ML rapidly declines to near-depletion during the peak period of the diatom bloom in spring, which results in an increasing consumption ratio of silicate to nitrogenous nutrients by diatoms as they become more iron-limited, causing a more rapid decrease of silicate compared to that of nitrogenous nutrients in the ML, followed by the silicate limitation of diatoms, and (3) Fed supplied to the ML by dust dissolution and desorption from particulate iron, by alleviating iron limitation of phytoplankton, supports their continuous utilization of nitrate from spring to fall even though Fed concentration in the ML remains low after the peak spring bloom. The model explained quantitatively the above behavior of Fed and other nutrients associated with Fed over the annual cycle in the Oyashio region. © 2012 American Geophysical Union. All Rights Reserved.