The weak-coupling expansion for the pressure of thermal field theories does not converge unless the temperature is many orders of magnitude higher than those relevant for experiment at RHIC and CERN. In this talk, I will discuss screened perturbation theory (SPT), which is reorganization of the perturbative expansion. The pressure is computed to three loops and it is shown that the convergence improves dramatically. The generalization to gauge theories, which is called hard-thermal-loop perturbation theory (HTL), is briefly discussed.