The linear relation between exercise intensity and the increase rate of the neuromuscular activity assessed by electromyography allows the estimation of the fatigue threshold, which would be the intensity that could be maintained indefinitely without presenting increase on the electromyography signals along time. It has been hypothesized that the subjetive perception of exertion would behave similarly to the neuromuscular activity and that a perceived exertion threshold (PET) identified similarly to the electromyographic fatigue threshold could coincide with the critical velocity (CritV). Thirteen individuals from both genders (23.0 ± 2.5 years), in a 15 m long x 2.5 m deep swimming pool performed three deep water running exhaustive tests for the determination of the parameters of the critical velocity model, reporting the perceived exertion (6-20 points in Borg scale) each 15 m. For the PET identification, the straight lines inclination coefficients of the increase on the perceived exertion in time (ordinate) and the velocities used (abscissa) were adjusted into a linear function that provided a point in the velocity axis where, theoretically, the perceived exertion would be indefinitely stable. The CritV was estimated through the equations used in the critical velocity model. For comparison purposes of the CritV and PET estimations and their associations, the repeated measures analysis of variance ANOVA was used (p < 0.05) and the Pearson correlation was calculated. The data obtained for the CritV determination fulfilled the criteria adopted for the model's validity and CritV and PET presented no statistical difference (0.23 ± 0.02 m/s x 0.24 ± 0.03 m/s), being significantly correlated (r = 0.85). These results suggest that the PET seems to represent the maximum exercise intensity in which the physiological and psychophysical variables would attain stability and that this index may be used in the CritV determination.