Atmospheric heating gradients caused by upper tropospheric (UT) clouds, most abundant in the tropics, influence the atmospheric circulation which then affects patterns of precipitation. We use climate simulations to investigate how the radiative heating from UT clouds affects the large-scale circulation and how the circulation, in turn, influences precipitation. First, we present new observational metrics derived from satellite observations to demonstrate how much the simulated UT cloud properties (ice water path and radiative heating rates) depend on the parameterizations of UT cloud ice and microphysics (single-scattering properties and fall speed). We then examine the consequences of UT cloud – radiation interactions on the Hadley circulation by performing two sets of experiments with UT clouds transparent to radiation, differentiating between semi-transparent and opaque UT clouds. Using ensemble simulations, we are also able to separate immediate effects and feedbacks which build up subsequently. The distribution of UT clouds with respect to their optical depth is key in the sign of the dynamical effect: Semi-transparent cirrus weaken the Hadley circulation by decreasing the meridional heating gradients, whereas opaque UT clouds strengthen the circulation by concentrating radiative heating in the core of the Intertropical Convergence Zone. Furthermore, the radiative heating of the opaque UT clouds is strong enough to considerably affect the precipitation through changes in moisture convergence which itself has been changed by the Hadley circulation.