The thermal fatigue problem associated with the operation of supersonic transport aircraft has been studied, both analytically and experimentally, by mechanical simulation of the thermal stress conditions, with particular respect to the effect of displacement controlled buckling on the fatigue life of simple notched sheet coupon specimens. The design and manufacture of a special fatigue testing machine is described, in which the sheet specimens were subjected to constant amplitude displacement controlled cycling over a range of mean displacement, including various degrees of buckling, A comprehensive programme of tests on buckling and non-buckling strut specimens, at both room temperature and a constant elevated temperature of 150°C, is reported. A simple analysis of the displacement controlled buckling of a slender encastre strut has been satisfactorily developed, and the effects of initial curvature and end alignment errors are shown to alleviate the peak stress conditions applicable to an initially straight perfectly aligned strut. Using the stress-displacement relationships derived from this theoretical buckling analysis, which have been correlated with experimentally measured stresses, two methods of fatigue life prediction are compared, and the results are correlated with the fatigue lives achieved in the experimental investigation. From this correlation, three main conclusions are drawn: (i) Accurately determined surface stresses are essential in order to achieve realistic fatigue life predictions for a buckling strut, (ii) Cyclic buckling does not of itself introduce any significant life reduction phenomenon, (iii) The mode of failure of a buckling strut changes as the cyclic buckling is made more significant, and it tends towards a finite fatigue life with increasing compression. Examination of related work on thermal fatigue testing demonstrates the damaging influence of cyclic temperatures and the differences between thermal stress and mechanical stress fatigue.