The goal of this study is to study the in?uence of phytoplankton on the tropical Paci?c climate. An intermediate coupled model for the tropical Paci?c is used, which is based on the primitive equations and reduced gravity approximations and formulated on sigma-coordinates underneath a dynamic mixed layer. It is driven by an atmospheric mixed layer model combined with statistical wind-stress forcing. This statistical atmosphere relates anomalies from the mean temperature to wind-stress anomalies, amplifying the annual cycle and allowing for interannual variability through the Bjerknes feedback. Ocean biology is simulated using a nine-component NPDZ ecosystem model, which is coupled to the physical ocean model. The simulated chlorophyll concentrations feed back to the ocean dynamics by modifying the absorption of photosynthetically active radiation (PAR) in the surface layers, thereby changing the heating pro?le. It is shown that the simulated mean state, annual cycle and interannual anomalies associated with the ENSO phenomenon are signi?cantly in?uenced by the spatiotemporal characteristics of chlorophyll. This study extends the study of Timmermann and Jin [2003], which predicts an in?uence of the tropical Paci?c climate mean state and its variability. The feedback discussed here works as follows: Maxima of the chlorophyll concentration close to the base of the mixed layer lead to a positive buoyancy forcing within the surface layer. The resulting deepening of the eastern equatorial mixed layer leads to a reduction of the surface ocean currents. The south equatorial current, weakened by this e?ect generates an eastern Paci?c warming. Finally this process is enhanced by the positive Bjerknes feedback. Due to the deepening of the mixed layer, also the strength of the annual cycle is reduced. As a result of weakened nonlinear coupling to the annual cycle, ENSO variability is increased.