The dryness of terrestrial climate can be measured by the ratio of annual precipitation (P) to potential evapotranspiration (PET), where the latter represents the evaporative demand of the atmosphere, which depends on the surface air temperature, relative humidity, wind speed, and available energy. This study examines how the terrestrial mean aridity responds to global warming in terms of P/PET using the Coupled Model Intercomparison Project phase 5 transient CO 2 increase to 2 × CO 2 simulations. We show that the (percentage) increase (rate) in P averaged over land is~1.7%/°C ocean mean surface air temperature increase, while the increase in PET is 5.3%/°C, leading to a decrease in P/PET (i.e., a drier terrestrial climate) by~3.4%/°C. Noting a similar rate of percentage increase in P over land to that in evaporation (E) over ocean, we propose a framework for examining the change in P/PET, in which we compare the change in PET over land and E over ocean, both expressed using the Penman-Monteith formula. We show that a drier terrestrial climate is caused by (i) enhanced land warming relative to the ocean, (ii) a decrease in relative humidity over land but an increase over ocean, (iii) part of increase in net downward surface radiation going into the deep ocean, and (iv) different responses of PET over land and E over ocean for given changes in atmospheric conditions (largely associated with changes in temperatures). The relative contributions to the change in terrestrial mean aridity from these four factors are about 35%, 35%, 15%, and 15%, respectively. The slight slowdown of the surface wind over both land and ocean has little impact on the terrestrial mean aridity.