Self‐talk is an effective mental training technique that has been shown to facilitate or debilitate an athlete's performance, depending on its valence. Although the effects of self‐talk have been supported by observing change in sport performance, little is known about how self‐talk can induce physiological changes. Specifically, it is important to understand if the type of self‐talk (positive, neutral, or negative) and can influence stress‐related parameters, such as perceived exertion, cardiorespiratory, and cortisol responses. The study's objective was therefore to investigate the top‐down effect of positive and negative self‐talk compared to a dissociative activity during an iso‐metabolic running exercise on autonomic regulation of cardiorespiratory function. Twenty‐nine well‐trained male runners [38 ± 13 years, 177 ± 7 cm and 73 ± 7 kg] volunteered to participate in a randomized‐group design study that included a negative self‐talk (NST), a positive self‐talk, and a dissociative group (DG). First, participants underwent an incremental running test on a treadmill to determine the maximal oxygen uptake (V̇O2max). Next, participants received a mental training session on self‐talk and created three positive and three negative self‐talk statements. Finally, participants underwent a 60‐min steady‐state running exercise on a treadmill at 70% of V̇O2max, during which they were cued at 20‐, 35‐, and 50‐min with their personal self‐created positive or negative self‐talk statements while the DG listened to a documentary. Cardiorespiratory parameters and rate of perceived exertion (RPE) were recorded throughout the 60‐min endurance exercise. In addition, salivary cortisol samples were obtained at waking and after treatment. Although oxygen uptake, carbon dioxide production, RPE, and heart rate significantly changed overtime during the 60‐min steady‐state running exercise, no significant main treatment effect was found. However, RPE scores, minute ventilation, breathing frequency, and salivary cortisol were significantly higher in the NST group compared to the two other groups. These data suggest that NST [emotion‐induced stress, as reflected by elevated cortisol] altered the breathing frequency response. In conclusion, manipulating self‐talk alters hormonal response patterns, modulates cardiorespiratory function, and influences perceived exertion.