Background: Critical speed (CS) represents the asymptote of the speed-time relationship and is considered an estimate of the highest metabolic steady-state intensity. It is an important physiological parameter for training prescription in swimmers. Traditional determination requires multiple maximal trials, increasing testing time and athlete burden. The 3-minute all-out test (3MT) has been proposed as a practical single-session alternative to estimate CS. Aim: To evaluate whether the CS obtained from 100, 200, and 400 m performances corresponds to the CS derived from the 3MT. Methods: Fifteen adolescent competitive swimmers (16.0 ± 2.3 years; 6 males and 9 females) completed four testing sessions separated by 48 hours. Protocols included the 3MT and maximal 100, 200, and 400 m swimming tests. Time was recorded using electronic stopwatches and confirmed by video. CS was calculated as: (1) CS3MT, the mean speed during the final 30 s of the 3MT; and (2) the linear regression method using 100 m (72.0 ± 7.8 s), 200 m (163.8 ± 18.0 s), and 400 m (363.6 ± 37.2 s) performances. Statistical analyses included the Two One-Sided Test (TOST), Intraclass Correlation Coefficient (ICC), and Bland-Altman analysis. Results: CS3MT was 1.09 ± 0.11 m·s*¹, whereas the linear regression method yielded 1.04 ± 0.12 m·s*¹. TOST indicated that the methods were not statistically equivalent (p = 0.44). The ICC was 0.46 (p = 0.026), indicating modest correlation. Bland-Altman analysis showed limits of agreement from *0.30 to +0.19 m·s*¹ and a systematic bias of *0.057 m·s*¹. Conclusions: CS derived from the 3MT was not equivalent to CS obtained from the distance-time model. The modest correlation and wide limits of agreement suggest caution when using these methods interchangeably for training prescription. CS3MT tended to provide slightly higher values than the traditional linear regression method.
© Copyright 2026 XVth International Symposium on Biomechanics and Medicine in Swimming Proceedings. All rights reserved.