A method and apparatus is disclosed for cooling a substrate between high temperature thermal processing steps. In the disclosed embodiment, one or more cooling stations are located off-line within a wafer handling chamber, just outside the thermal processing chamber. After thermal processing, a hot wafer can be loaded on to the cooling station, where the wafer is subjected to forced convection cooling. In particular, the wafer is subjected to cooling gas from above and below through perforated upper and lower shower head assemblies. The wafer can thus be cooled rapidly on a cooling station while other wafers are transferred into and out of the processing chamber. Desirably, the wafer is cooled on the cooling station to a point at which it can be handled by a low temperature wafer handler and stored in a low temperature cassette.