This study aimed to evaluate if the changes in oxygen saturation levels at intercostal muscles (SmO₂-m.intercostales) assessed by near-infrared spectroscopy (NIRS) using a wearable device could determine the respiratory compensation point (RCP) during exercise. Fifteen healthy competitive triathletes (eight males; 29 ± 6 years; height 167.6 ± 25.6 cm; weight 69.2 ± 9.4 kg; $\dot{\mathrm{V}}$O₂-máx 58.4 ± 8.1 mL·kg⁻¹·min⁻¹) were evaluated in a cycle ergometer during the maximal oxygen-uptake test ($\dot{\mathrm{V}}$O₂-máx), while lung ventilation ($\dot{\mathrm{V}}$E), power output (watts, W) and SmO₂-m.intercostales were measured. RCP was determined by visual method (RCP_visual: changes at ventilatory equivalents ($\dot{\mathrm{V}}$E·$\dot{\mathrm{V}}$CO₂⁻¹, $\dot{\mathrm{V}}$E·$\dot{\mathrm{V}}$O₂⁻¹) and end-tidal respiratory pressure (PetO₂, PetCO₂) and NIRS method (RCP_NIRS: breakpoint of fall in SmO₂-m.intercostales). During exercise, SmO₂-m.intercostales decreased continuously showing a higher decrease when $\dot{\mathrm{V}}$E increased abruptly. A good agreement between methods used to determine RCP was found (visual vs NIRS) at %$\dot{\mathrm{V}}$O₂-máx, $\dot{\mathrm{V}}$O₂, $\dot{\mathrm{V}}$E, and W (Bland-Altman test). Correlations were found to each parameters analyzed (r = 0.854; r = 0.865; r = 0.981; and r = 0,968; respectively. p < 0.001 in all variables, Pearson test), with no differences (p < 0.001 in all variables, Student’s t-test) between methods used (RCP_visual and RCP_NIRS). We concluded that changes at SmO₂-m.intercostales measured by NIRS could adequately determine RCP in triathletes.