The ratio of liquid water to ice in a cloud, largely controlled by the presence of ice nuclei and cloud temperature, alters cloud radiative effects. This study quantitatively examines how the liquid fraction of clouds influences various climate feedbacks using the NCAR Community Atmosphere Model (CAM). Climate feedback parameters were calculated using equilibrated temperature changes in response to increases in the atmospheric concentration of carbon dioxide in CAM Version 3.0 with a slab ocean model. Two sets of model experiments are designed such that cloud liquid fraction linearly decreases with a decrease in temperature down to À20°C (Experiment "C20") and À40°C (Experiment "C40"). Thus, at the same subzero temperature, C20 yields fewer liquid droplets (and more ice crystals) than C40. Comparison of the results of experiments C20 and C40 reveals that experiment C20 is characterized by stronger cloud and temperature feedbacks in the tropics (30°N-30°S) (by 0.25 and À0.28 W m À2 K
À1, respectively) but weaker cloud, temperature, and albedo feedbacks (by À0.20, 0.11, and À0.07 W m À2 K
À1) in the extratropics. Compensation of these climate feedback changes leads to a net climate feedback change of~7.28% of that of C40 in the model. These results suggest that adjustment of the cloud phase function affects all types of feedbacks (with the smallest effect on water vapor feedback). Although the net change in total climate feedback is small due to the cancellation of positive and negative individual feedback changes, some of the individual changes are relatively large. This illustrates the importance of the influence of cloud phase partitioning for all major climate feedbacks, and by extension, for future climate change predictions.