Jupiter’s upper troposphere and stratosphere are host to a rich dynamical and chemical activity. This modulates the thermal structure and distribution of trace species and aerosols, which, in turn, impact the atmospheric radiative budget and dynamics. In this paper, we present a computationally efficient 1-D seasonal radiative model, with convective adjustment, of Jupiter’s atmosphere. Our model takes into account radiative forcings from the main hydrocarbons (methane, ethane, acetylene), ammonia, collision-induced absorption, several cloud and haze layers and an internal heat flux. We parametrize four tropospheric cloud and haze layers. Three of them (one tropospheric cloud near 800 mbar, one upper tropospheric haze, one stratospheric haze) are set to be uniform with latitude. On the contrary, we prescribe the spatial distribution of another UV-absorbing “polar” stratospheric haze comprising fractal aggregates based on published observational constraints, as their concentration varies significantly with latitude. We detail sensitivity studies of the equilibrium temperature profile to several parameters (hydrocarbon abundances, cloud particle sizes and optical depths, optical properties of the stratospheric polar haze, etc.). We then discuss the expected seasonal, vertical and meridional temperature variations in Jupiter’s atmosphere and compare the modeled thermal structure to that derived from Cassini and ground-based thermal infrared observations.We find that the equilibrium temperature in the 5–30 mbar pressure range is very sensitive to the chosen stratospheric haze optical properties, sizes and number of monomers. One of the three sets of optical properties tested yields equilibrium temperatures that match well, to first order, the observed ones. In this scenario, the polar haze significantly warms the lower stratosphere (10–30 mbar) by up to 20 K at latitudes 45–60°, and reproduces an observed north–south asymmetry in stratospheric temperature. The polar haze also acts to shorten significantly the radiative timescales, estimated by our model to 100 (Earth) days at the 10-mbar level. At pressures lower than 3 mbar, our modeled temperatures systematically underestimate the observed ones by ∼5 K. This might suggest that other processes, such as dynamical heating by wave breaking or by eddies, or a coupling with thermospheric circulation, play an important role. An alternate possibility is that the uncertainty on the abundance of hydrocarbons is responsible for this mismatch. In the troposphere, we can only match the observed lack of meridional gradient of temperature by varying the internal heat flux with latitude.We then exploit knowledge of heating and cooling rates (using our radiative seasonal model combined to observational constraints on the temperature) to diagnose the residual-mean circulation in Jupiter’s stratosphere. This is done under the assumption that the eddy heat flux convergence term is negligible. In the Earth’s stratosphere, the residual-mean circulation obtained with this method represents well, on a seasonal scale, the transport of tracers in regions where wave breaking and dissipation are weak. However, on Jupiter, in the lower stratosphere (5–30 mbar), the residual-mean circulation strongly depends on the assumed properties of the stratospheric haze. Our main conclusion is that it is crucial to improve our knowledge on the different radiative forcing terms (in particular regarding the stratospheric haze properties) to increase our confidence in the estimated circulation. By extension, this will also be crucial for future 3D GCM studies.