The study reports on the direct aerosol effect (DAE) radiative forcing (RF) from AeroCom Phase II simulations, where 16 global aerosol models were used to estimate the RF of anthropogenic aerosols. The models considered anthropogenic sulphate, black carbon (BC), and organic aerosols (OA), as well as nitrate and secondary organic aerosols (SOA). The simulated all-sky RF of the DAE from total anthropogenic aerosols ranged from -0.58 to -0.02 Wm⁻², with a mean of -0.27 Wm⁻². Accounting for missing aerosol components reduced the range and slightly strengthened the mean. Modifying the model estimates for missing aerosol components and the time period 1750–2010 resulted in a mean RF of -0.35 Wm⁻². Compared to AeroCom Phase I, the DAE RF is more negative, and the RF from BC from fossil fuel and biofuel emissions is stronger positive. Models with strong positive BC RF also tend to have strong negative sulphate or OA RF, leading to smaller uncertainty in the total DAE RF. The spread in results for individual aerosol components is substantial, influenced by differences in burden, mass extinction coefficient (MEC), and normalized RF with respect to AOD. The mean RF of the total DAE from 16 models was -0.27 Wm⁻², stronger negative than the -0.22 Wm⁻² from AeroCom Phase I. The RF of sulphate was -0.32 Wm⁻², weaker than the -0.35 Wm⁻² from Phase I. The RF of BC was positive, with a mean of 0.18 Wm⁻². The RF of OA was weak, with a mean of -0.03 Wm⁻². The RF of SOA ranged from -0.08 to -0.02 Wm⁻². The RF of nitrate was -0.08 Wm⁻². The RF of biomass burning (BB) aerosols varied, with a mean close to zero. The study highlights the importance of considering aerosol components and their interactions in estimating the DAE RF. The results show that the total DAE RF is less uncertain than the sum of individual aerosol components due to compensating effects between different aerosol types. The study also emphasizes the need for further research to understand model differences and improve the accuracy of DAE RF estimates.The study reports on the direct aerosol effect (DAE) radiative forcing (RF) from AeroCom Phase II simulations, where 16 global aerosol models were used to estimate the RF of anthropogenic aerosols. The models considered anthropogenic sulphate, black carbon (BC), and organic aerosols (OA), as well as nitrate and secondary organic aerosols (SOA). The simulated all-sky RF of the DAE from total anthropogenic aerosols ranged from -0.58 to -0.02 Wm⁻², with a mean of -0.27 Wm⁻². Accounting for missing aerosol components reduced the range and slightly strengthened the mean. Modifying the model estimates for missing aerosol components and the time period 1750–2010 resulted in a mean RF of -0.35 Wm⁻². Compared to AeroCom Phase I, the DAE RF is more negative, and the RF from BC from fossil fuel and biofuel emissions is stronger positive. Models with strong positive BC RF also tend to have strong negative sulphate or OA RF, leading to smaller uncertainty in the total DAE RF. The spread in results for individual aerosol components is substantial, influenced by differences in burden, mass extinction coefficient (MEC), and normalized RF with respect to AOD. The mean RF of the total DAE from 16 models was -0.27 Wm⁻², stronger negative than the -0.22 Wm⁻² from AeroCom Phase I. The RF of sulphate was -0.32 Wm⁻², weaker than the -0.35 Wm⁻² from Phase I. The RF of BC was positive, with a mean of 0.18 Wm⁻². The RF of OA was weak, with a mean of -0.03 Wm⁻². The RF of SOA ranged from -0.08 to -0.02 Wm⁻². The RF of nitrate was -0.08 Wm⁻². The RF of biomass burning (BB) aerosols varied, with a mean close to zero. The study highlights the importance of considering aerosol components and their interactions in estimating the DAE RF. The results show that the total DAE RF is less uncertain than the sum of individual aerosol components due to compensating effects between different aerosol types. The study also emphasizes the need for further research to understand model differences and improve the accuracy of DAE RF estimates.