Abstract. A primary goal of the Atmospheric Chemistry and Climate Model Intercomparison Project (ACCMIP) was to characterize the short-lived drivers of preindustrial to 2100 climate change in the current generation of climate models. Here we evaluate historical and future radiative forcing in the 10 ACCMIP models that included aerosols, 8 of which also participated in the Coupled Model Intercomparison Project phase 5 (CMIP5). The models generally reproduce present-day climatological total aerosol optical depth (AOD) relatively well. They have quite different contributions from various aerosol components to this total, however, and most appear to underestimate AOD over East Asia. The models generally capture 1980–2000 AOD trends fairly well, though they underpredict AOD increases over the Yellow/Eastern Sea. They appear to strongly underestimate absorbing AOD, especially in East Asia, South and Southeast Asia, South America and Southern Hemisphere Africa. We examined both the conventional direct radiative forcing at the tropopause (RF) and the forcing including rapid adjustments (adjusted forcing; AF, including direct and indirect effects). The models' calculated all aerosol all-sky 1850 to 2000 global mean annual average RF ranges from −0.06 to −0.49 W m−2, with a mean of −0.26 W m−2 and a median of −0.27 W m−2. Adjusting for missing aerosol components in some models brings the range to −0.12 to −0.62 W m−2, with a mean of −0.39 W m−2. Screening the models based on their ability to capture spatial patterns and magnitudes of AOD and AOD trends yields a quality-controlled mean of −0.42 W m−2 and range of −0.33 to −0.50 W m−2 (accounting for missing components). The CMIP5 subset of ACCMIP models spans −0.06 to −0.49 W m−2, suggesting some CMIP5 simulations likely have too little aerosol RF. A substantial, but not well quantified, contribution to historical aerosol RF may come from climate feedbacks (35 to −58 %). The mean aerosol AF during this period is −1.12 W m−2 (median value −1.16 W m−2, range −0.72 to −1.44 W m−2), indicating that adjustments to aerosols, which include cloud, water vapor and temperature, lead to stronger forcing than the aerosol direct RF. Both negative aerosol RF and AF are greatest over and near Europe, South and East Asia and North America during 1850 to 2000. AF, however, is positive over both polar regions, the Sahara, and the Karakoram. Annual average AF is stronger than 0.5 W m−2 over parts of the Arctic and more than 1.5 W m−2 during boreal summer. Examination of the regional pattern of RF and AF shows that the multi-model spread relative to the mean of AF is typically the same or smaller than that for RF over areas with substantial forcing. Historical aerosol RF peaks in nearly all models around 1980, declining thereafter. Aerosol RF declines greatly in most models over the 21st century and is only weakly sensitive to the particular Representative Concentration Pathway (RCP). One model, however, shows approximate stabilization at current RF levels under RCP 8.5, while two others show increasingly negative RF due to the influence of nitrate aerosols (which are not included in most models). Aerosol AF, in contrast, continues to become more negative during 1980 to 2000 despite the turnaround in RF. Total anthropogenic composition forcing (RF due to well-mixed greenhouse gases (WMGHGs) and ozone plus aerosol AF) shows substantial masking of greenhouse forcing by aerosols towards the end of the 20{th} century and in the early 21st century at the global scale. Regionally, net forcing is negative over most industrialized and biomass burning regions through 1980, but remains strongly negative only over East and Southeast Asia by 2000 and only over a very small part of Southeast Asia by 2030 (under RCP8.5). Net forcing is strongly positive by 1980 over the Sahara, Arabian peninsula, the Arctic, Southern Hemisphere South America, Australia and most of the oceans. Both the magnitude of and area covered by positive forcing expand steadily thereafter. There is no clear relationship between aerosol AF and climate sensitivity in the CMIP5 subset of ACCMIP models. There is a clear link between the strength of aerosol+ozone forcing and the global mean historical climate response to anthropogenic non-WMGHG forcing (ANWF). The models show ~20–35% greater climate sensitivity to ANWF than to WMGHG forcing, at least in part due to geographic differences in climate sensitivity. These lead to ~50% more warming in the Northern Hemisphere in response to increasing WMGHGs. This interhemispheric asymmetry is enhanced for ANWF by an additional 10–30%. At smaller spatial scales, response to ANWF and WMGHGs show distinct differences.