The present study aimed to investigate the effect of reflective self-talk interventions on athletes’ organic self-talk, with an emphasis on the moderating role of need for cognition. This study employed a quasi-experimental design with a pretest–posttest control group. The statistical population consisted of male athletes aged 18–25 years who regularly participated in sports activities. Based on inclusion criteria and levels of need for cognition, 60 athletes were purposively selected and categorized into high and low need for cognition groups. Participants within each level were then randomly assigned to either intervention or control groups. The intervention group underwent a structured reflective self-talk program over eight sessions across four weeks, while the control group continued their regular training routines. Research instruments included the Organic Self-Talk in Sport Questionnaire, the Need for Cognition Scale, and a cycling endurance performance test to exhaustion. Results of the 2×2 factorial analysis of covariance (ANCOVA) revealed that the reflective self-talk intervention significantly increased goal-directed organic self-talk and positive automatic organic self-talk while significantly reducing negative automatic organic self-talk (p < .001). Furthermore, compared with athletes with low need for cognition, athletes with high need for cognition demonstrated greater increases in goal-directed self-talk and greater reductions in negative self-talk following the intervention. However, the interaction effect of need for cognition on positive automatic self-talk was not significant. Overall, the findings suggest that reflective self-talk intervention may serve as an effective approach for reconstructing athletes’ inner speech, particularly among individuals with a high need for cognition.