Thermography is a non-invasive method for measuring surface temperatures and may be a convenient way of identifying hypo/hyperthermic areas under a saddle that may be related to saddle pressures. A thermal camera quantified minimum/maximum/mean temperatures at specific locations (left/right) of the thoracic region at three-time points: (1) baseline; (2) post lunging; (3) post ridden exercise in eight non-lame sports horses ridden by the same rider. A Pliance (Novel) pressure mat determined the mean/peak saddle pressures (kPa) in the cranial and caudal regions. General linear mixed models with the horse as the random factor investigated the time point (fixed factor: baseline; lunge; ridden) and saddle fit (fixed factor: correct; wide; narrow) on thermal parameters with Bonferroni post hoc comparison. The saddle pressure data (grouped: saddle width) were assessed with an ANOVA and Tukey post hoc comparison (p ≤ 0.05). Differences between the saddle widths in the cranial/caudal mean (p = 0.05) and peak saddle pressures (p = 0.01) were found. The maximum temperatures increased post lunge (p ≤ 0.0001) and post ridden (p ≤ 0.0001) compared to the baseline. No difference between post lunge and post ridden exercise (all p ≥ 0.51) was found. The thermal activity does not appear to be representative of increased saddle pressure values. The sole use of thermal imaging for saddle fitting should be applied with caution.