In this study, circulating β-endorphin, ACTH and cortisol concentrations were chosen as physiological markers of the hypothalamic-pituitary-adrenal (HPA) stress-related response in Therapeutic Riding Horses (TRHs). The aim of this research was to investigate HPA axis hormonal response in TRHs and check the effects of different kind of riders (mentally-impaired and able-minded riders). The main hypothesis was that, in TRHs, the HPA axis reaction would be less responsive to mentally-impaired than to able-minded riders, due to increasing levels of horse control by the able-minded riders. 6 TRHs were randomly ridden in the same setting by 2 groups of riders, adult mentally-impaired riders (Treatment Group A) and able-minded adult riders (Treatment Group B). Both groups consisted of inexperienced riders. Each rider rode 2 times a week for 6 consecutive weeks. The horses performed the same gaits and exercises at all sessions, with both Treatment Group A and Treatment Group B riders. All hormones’ concentrations were determined on a resting day (baseline, T0), before the session (T1), and at 5 min (T2) and 30 min (T3) after each session. Statistical analysis showed significant effects of the different Treatment Groups on the cortisol concentrations only, with lower cortisol concentrations both at T2 and at T3 in Treatment Group A than in Treatment Group B. Related to changes over the sampling time, ccortisol concentrations in horses were significantly higher at T1 than at T0 in Treatment Group A, and at T1, T2 and T3 than at T0, in Treatment Group B. ACTH concentrations were significantly higher at T1, T2 and T3 than at T0, only in Treatment Group B. β-endorphin concentrations at T3 were significantly higher than at T0, only in Treatment Group B. Results showed differences in the TRHs' stress levels between rider groups, with significantly lower cortisol concentrations in Treatment Group A than in Treatment Group B. Thus, although sessions’ workload stress was equivalent, the minor response of HPA axis could be related to different effects of horse-human interaction.