The Ridden Horse Pain Ethogram (RHpE) was developed to facilitate the identification of musculoskeletal pain. The aim of the current study was to collate behavioural data using the RHpE from horses at competitions assumed by their owners and/or riders to be fit for competition. The objectives were to quantify the frequency of occurrence of behaviours in pain-free horses and those with lameness or abnormalities of canter and to determine any differences between disciplines and levels of competition. The RHpE was applied to 1358 horses competing in Grand Prix (GP) dressage (n = 211), 5* three-day events (TDE) (n = 137), or low-level one-day events (ODE) (n = 1010). The median RHpE score for all horses was 4 (interquartile range [IQR] 2, 5; range 0, 12) and the median lameness grade was 0 (IQR 0, 1; range 0, 4). The Kruskal–Wallis test, followed by Dunn’s test for pairwise comparisons, found a difference in median RHpE scores between low-level ODE and GP dressage (p = 0.001), but not between 5* TDE and low-level ODE (p = 0.09) or between GP dressage and 5* TDE (p = 1.00). The median RHpE score was highest for low-level ODE. The Chi-square/Fisher’s exact test identified a significant difference in prevalence of most of the 24 behaviours of the RHpE in non-lame compared with lame horses. Recognition of the behaviours of non-lame horses may improve equine welfare and performance, and rider comfort, confidence, and safety.