There are many reports of land surface temperature (LST) anomalies appearing prior to large earthquakes. A number of methods have been applied in hindcast mode to identify these anomalies, using infrared datasets collected from Earth-orbiting remote sensing satellites. Here we examine three such methods and apply them to six years (2001–2006) of MODIS LST data collected over the region of the 2001 Gujarat (India) earthquake, which previous studies have identified as a site exhibiting possible pre-seismic and post-seismic thermal anomalies. Methods 1 and 2 use an LST differencing technique, while Method 3, the Robust Satellite Technique (RST), has been developed specifically for the identification of thermal anomalies within spatio-temporal datasets. In relation to the Gujarat Earthquake, results from Methods 1 and 2 (LST differencing) indicate that changes previously reported to be potential precursory thermal ‘anomalies’ appear instead to occur within the range of normal thermal variability. Results obtained with Method 3 (RST) do appear to show significant ‘anomalies’ around the time of the earthquake, but we find these to be related to positive biases caused by the presence of MODIS LST data gaps, attributable to cloud cover and mosaicing of neighboring orbits of data. Currently, therefore, we find no convincing evidence of LST precursors to the 2001 Gujarat earthquake, and urge care in the use of approaches aimed at identifying such seismic thermal anomalies.