This study assesses climate feedbacks in coupled ocean–atmosphere models using twenty-first-century climate change experiments. Water vapor is found to provide the largest positive feedback in all models, with its strength consistent with constant relative humidity changes. Cloud and surface albedo feedbacks are also positive, while the only stabilizing feedback comes from the temperature response. Large intermodel differences in the lapse rate feedback are observed, associated with regional surface warming patterns. The sum of lapse rate and water vapor feedbacks is small across models, but cloud feedbacks are the largest source of uncertainty in climate sensitivity predictions. The study uses a consistent methodology to compare feedback strengths in 14 coupled ocean–atmosphere models. Feedbacks are calculated based on changes in global mean surface temperature and radiative flux at the top of the atmosphere. Water vapor feedback is found to be the strongest, followed by cloud and surface albedo feedbacks. The troposphere warms faster than the surface, resulting in a negative lapse rate feedback. Intermodel differences in cloud feedback are the largest contributor to uncertainty in climate sensitivity. The study also finds that the range of water vapor and lapse rate feedbacks in the IPCC AR4 models is smaller than previously reported, likely due to differences in feedback methodologies. However, the combined lapse rate and water vapor feedback is more consistent across models. The results suggest that cloud feedback is the largest uncertainty in climate sensitivity predictions, with a standard deviation of 0.37, significantly larger than other feedbacks. The study concludes that water vapor provides the largest positive feedback, and cloud feedback is the largest source of uncertainty in climate sensitivity predictions. The results highlight the importance of accurately assessing feedback strengths in climate models to improve predictions of climate sensitivity. The methodology used in this study provides a consistent and economical way to compare feedbacks among different models.This study assesses climate feedbacks in coupled ocean–atmosphere models using twenty-first-century climate change experiments. Water vapor is found to provide the largest positive feedback in all models, with its strength consistent with constant relative humidity changes. Cloud and surface albedo feedbacks are also positive, while the only stabilizing feedback comes from the temperature response. Large intermodel differences in the lapse rate feedback are observed, associated with regional surface warming patterns. The sum of lapse rate and water vapor feedbacks is small across models, but cloud feedbacks are the largest source of uncertainty in climate sensitivity predictions. The study uses a consistent methodology to compare feedback strengths in 14 coupled ocean–atmosphere models. Feedbacks are calculated based on changes in global mean surface temperature and radiative flux at the top of the atmosphere. Water vapor feedback is found to be the strongest, followed by cloud and surface albedo feedbacks. The troposphere warms faster than the surface, resulting in a negative lapse rate feedback. Intermodel differences in cloud feedback are the largest contributor to uncertainty in climate sensitivity. The study also finds that the range of water vapor and lapse rate feedbacks in the IPCC AR4 models is smaller than previously reported, likely due to differences in feedback methodologies. However, the combined lapse rate and water vapor feedback is more consistent across models. The results suggest that cloud feedback is the largest uncertainty in climate sensitivity predictions, with a standard deviation of 0.37, significantly larger than other feedbacks. The study concludes that water vapor provides the largest positive feedback, and cloud feedback is the largest source of uncertainty in climate sensitivity predictions. The results highlight the importance of accurately assessing feedback strengths in climate models to improve predictions of climate sensitivity. The methodology used in this study provides a consistent and economical way to compare feedbacks among different models.