Mechanoluminescent (ML) materials have mechanical-light conversion properties and can generate luminescence under mechanical stress, which makes ML materials have high application value in optical information display. In this paper, the crystal structure and defect distribution were adjusted by changing the K<sup>+</sup>/Na<sup>+</sup> ratio of the ferroelectric matrix K<sub><i>x</i></sub>Na<sub>1-<i>x</i></sub>NbO<sub>3</sub>:0.5%Pr<sup>3+</sup>(K<sub><i>x</i></sub>NNOP), the effects of K<sup>+</sup> content on the photoluminescence (PL) and ML properties were systematically investigated. The research results indicated that as the K<sup>+</sup> content increased, the symmetry of the crystal enhanced, resulting in a decrease in the PL intensity of the K<sub><i>x</i></sub>NNOP samples. It is worth noting that the emission peaks caused by the <sup>3</sup>P<sub>1</sub>→<sup>3</sup>H<sub>5</sub>and <sup>3</sup>P<sub>0</sub>→<sup>3</sup>H<sub>5</sub> transition at the Pr<sup>3+</sup> electron level appeared in the PL spectra of the components with higher K<sup>+</sup> content under the light excitation of 450 nm, which was attributed to the different energy level positions of the internal valence electron charge transfer states (IVCT) within Pr-O-Nb caused by the change in the distance between Pr<sup>3+</sup> and Nb<sup>5+</sup>. Under compressive stress, The K<sub><i>x</i></sub>NNOP (<i>x</i>=0, 0.01, 0.02, 0.1) components exhibited bright red ML, and the ML intensity increased with the increase of K<sup>+</sup> content, the K<sub>0.1</sub>NNOP component exhibited the highest ML intensity emission. In particular, the ML behavior had the characteristics of repeatability and recoverability. The trap energy levels in the K<sub><i>x</i></sub>NNOP samples were investigated by thermoluminescence curves, revealing that the enhancement of ML in K<sub>0.1</sub>NNOP may be related to the differences in trap density and trap depth caused by changes in K<sup>+</sup> content. Based on these results, a model was established to elucidate the possible ML mechanism in K<sub><i>x</i></sub>NNOP.
