In this study, the response of the climate system to aerosol direct and indirect radiative forcing is investigated. Several multidecadal equilibrium simulations are carried out, using the NCAR CCM3 framework coupled to a separately developed aerosol treatment. The aerosol treatment includes, e.g., a life cycle scheme for particulate sulfate and black carbon (natural and anthropogenic), calculations of aerosol size distributions and CCN activation, as well as computations of direct and indirect forcing (1st and 2nd indirect effect). In all the simulations the full aerosol treatment is run online, hence responding interactively to changes in climate. By far the largest response is caused by the indirect forcing, with a globally averaged cooling of -1.25 K due to anthropogenic aerosols. The largest temperature reduction is found in the Northern Hemisphere because of a larger aerosol burden there. As a result of this cooling pattern, the Intertropical Convergence Zone is displaced southward by a few hundred kilometers. Interestingly, a similar, though less significant displacement is also found in the experiments with the direct effect alone, even though the globally averaged aerosol induced cooling is much weaker in that case, i.e., -0.08 K. The direct radiative forcing is much stronger at the surface than at the top of the atmosphere, and this leads to a slight weakening of the hydrological cycle. Comparing simulations with direct and indirect forcing combined to those with indirect and direct forcing separately, a residual, caused by nonlinear model feedbacks, is manifested through a reduction in precipitation. This reduction amounts to -0.5% in a global average and exceeds -2.5% in the Arctic, highlighting the role of high-latitude climate feedbacks. Globally, cloud feedback is negative in the sense that in the colder climate resulting from anthropogenic aerosol forcing, net cloud forcing is reduced by 15% compared to the original climate state. This is caused by a general cloud thinning, especially at high latitudes, while in the most polluted regions, clouds are thicker through the 2nd indirect effect. The 1st indirect effect, on the other hand, remains intact in the presence of climate feedbacks, yielding a similar signature of cloud droplet reduction as in the pure forcing simulations.