The purpose of this study was to compare heart rate variability threshold (HRVT) in 6 incremental tests and test its reproducibility using visual inspection and Dmax methods for root mean square of successive differences between the adjacent normal R-R intervals (RMSSD), standard deviation of the normal RR interval (SDNN) and standard deviation of instantaneous beat-to-beat variability (SD1). 12 adult males performed an incremental test to volitional fatigue on a cycle simulator during 6 visits to the laboratory. The initial test load was 25 W, and the intensity was increased by 25 W every 3 min until volitional fatigue set in. The HRV during the incremental test was analyzed using the RMSSD, SDNN and SD1 indices and the determination of HRVT was performed using 2 methods: visual inspection and Dmax. The results demonstrated that the SD1 and RMSSD indices, determined by the visual inspection method, presented the highest reproducibility of HRVT when compared with the other indices and methods. We concluded that the best method for determining HRVT was the technique using the point of stabilization by visual inspection in the SD1 and RMSSD indices during the incremental test, due to its high reproducibility, lower coefficient of variation and increment size.