Ultrasonic estimation of heat-induced echo strain has been suggested as a noninvasive technique for guiding focused ultrasound (US) surgery (FUS), that is, for predicting the location of the thermal lesion before it is formed. The proposed strategy is to run the FUS system at a nonablative intensity and to use a diagnostic transducer to image the heat-induced echo strain, which, over a sufficiently small temperature range, is proportional to the temperature rise. The principal aim of this in vitro study was to determine if temperature-induced strain imaging is likely to be able to visualise the small (< 0.5%) strains that one would be restricted to in vivo. Temperature rises ranging from approximately 2 degrees C to 15 degrees C (starting at approximately 25 degrees C) were induced in bovine liver samples using an FUS system. The pre- and post-heated US images were processed to produce images of the apparent axial strain. These images were found to possess excellent spatial and contrast resolution, so that the hot spot remained clearly visible even when the spatial peak strain value was approximately 0.2% (corresponding to temperature rises on the order of 2 to 5 degrees C). Good repeatability in the strain images was observed within and between tissue samples. Artefacts due to thermoacoustic refraction were seen distal to the heated region, but they did not reduce hot spot visibility. The length of the hot spot exceeded that of the subsequent ablation (by approximately 200%), which was to be expected given that temperature imaging depicts the entire area over which the temperature has increased relative to the baseline. We conclude that temperature-induced strain imaging for the guidance of FUS in the liver is likely to be feasible, provided that it will be possible either to neglect or to correct for the additional sources of error (such as cardiac-induced motion) that will arise in vivo.