The flash technique for measuring thermal diffusivity is analyzed for the case of a cylindrical‐shaped specimen of radiusr0and thicknessato determine the effects of radiation at high temperatures, finite duration of the heat pulse, and the feasibility of low temperature measurements. It is found that the flash diffusivity method is useful in two complementary limits: (1) pulse time &tgr; short compared to the characteristic thermal response timetc, (2) &tgr;/tcof the order 1 to 10. The former case corresponds to the original description of Parker, Jenkins, and Abbott, while the latter case is suitable at very low temperatures. Moreover, it is shown that there is an optimum specimen thicknessafor a given material and pulse time &tgr;, in the sense that a higher temperature can be reached before any corrections have to be made to the Parkeret al.analysis.