The objective of this research was to determine whether the behaviour of thermophilic bacteria growing as biofilms would change after being subjected to temperature cycling. It was used to test the practicality of a temperature cycling method for use in the heat transfer sections (milk pre-heaters and evaporators) of dairy plants for reducing the biofilm growth of thermophilic bacteria. Biofilms may protect bacterial cells from cleaning-in-place operations (CIP). CIP is the standard technique for cleaning dairy plant in New Zealand (AS/NZS 2541, 1998). If the temperature cycling method is used in the heat transfer sections of dairy plants, surviving thermophilic bacteria in the form of biofilms may be continuously exposed to the temperature cycling environment; thus a sub-population which is resistant to temperature cycling may adhere in greater numbers or grow more efficiently under the selective conditions. Wild-type strains of Geobacillus stearothermophilus and Anoxybacillus flavithermus were isolated from a commercial milk powder sample obtained from a milk powder production plant in New Zealand. These two wild-type strains were subjected to a temperature cycling regime consisting of 55ºC for 15 minutes then 35ºC for 35 minutes for 192 hours using a modified Polymerase Chain Reaction (PCR) thermocycler and hexagonal flat stainless steel reactors. At the end of each of four serial experiments, sub-population isolates for both species were obtained from the effluent milk. Wild-type strains and the sub-population strains of these two organisms were tested for the ability to adhere and grow as biofilms on stainless steel surfaces by using a Centre for Disease Control (CDC) biofilm reactor. The cell counts per unit area of adhesion, maximum specific growth rates (μmax) of the biofilms, and cells released in out-flowing milk were measured. The analysed data of the wild-type strain and the sub-population strain of each organism were compared by using p-value of 2 samples t-test statistical analysis to check whether the strains had adapted to the temperature cycling treatment. The results showed that the ability of planktonic cells of the wild-type A. flavithermus strain to adhere to stainless steel surfaces was significantly changed with p-value = 0.001 (< 0.05) after being subjected to 192 hours temperature cycling, but the sub-population growth as biofilms on stainless steel surfaces was no different from the wild-type strain with p-value = 0.235 (> 0.05). In contrast, the sub-population of G. stearothermophilus was significantly different from the wild-type after being subjected to 192 hours temperature cycling. The G. stearothermophilus sub-population strain was more resistant to temperature cycling, having a greater ability to adhere (p-value = 0.000 < 0.05) and grow as a biofilm (p-value = 0.008 < 0.05) than the wild-type strain. Since G. stearothermophilus is a common contaminant in dairy plants, the temperature cycling method may not be a viable long-term solution for dairy plants for reducing the growth rates of thermophilic bacteria, and further studies, such as different temperature profiles both in terms of magnitude and duration, are required.