THE standard description of the austenite–martensite transformation includes the statement that once the material reaches the martensite start temperature, or below, then the transformation occurs immediately1. Hence, if a nucleus is present in the material, then growth occurs to an extent determined by the transformation temperature. We expect, therefore, that there should be no time delay in this transformation and that the amount of martensite should depend on the transformation temperature, for a given grain size and composition2. If there is a time delay in the appearance of martensite, again at a temperature at or below the martensite start temperature (Ms), then it is important to show this by establishing that no growth of martensite occurs while a specimen is held at or below Ms. We show here that the nucleation event and not the growth behaviour of martensite can be affected by a common metallurgical treatment. The experiment is carried out at a constant grain size and composition with the result that we can unambiguously relate the change in nucleation rate to the present procedure.