Perceptual Demands in CrossFit®: Convergent Validity of sRPE and the Relative Role of Session Phases in a Prospective Observational Study

The purpose of this study was to examine the convergent validity and reliability of the session rating of perceived exertion (sRPE) in complete CrossFit® sessions, comparing it with a weighted RPE measure (RPEW) obtained from different session phases. Twenty-four recreational practitioners (13 men and 11 women) completed 28 standardized sessions of approximately 60 minutes, consisting of warm-up, strength/skill, WOD, and cooldown. RPE was collected after each phase, and sRPE was recorded 30 minutes post-session. Results showed that RPEW (5.8 ± 1.5) was significantly lower than sRPE (6.8 ± 1.4; p < 0.001, d = 0.69), resulting in a higher training load estimated by sRPE (+15.5%). Bland–Altman analysis revealed a positive bias and wide limits of agreement, while relative reliability was moderate-to-good (ICC = 0.73–0.77). Multiple regression analysis indicated that WOD RPE explained 70% of the variance in sRPE (R2 = 0.70, p < 0.001), confirming its role as the primary perceptual determinant. In conclusion, sRPE represents a valid, simple, and cost-effective tool for monitoring internal load in CrossFit®. However, it tends to overestimate weighted RPE and is strongly influenced by the final phase of the session, which should be taken into account by coaches and practitioners when interpreting its values.