This work aims at presenting a method for automatic detection of the anaerobic thresholds by surface electromyography (RMS-slope) and compare with respective values identified by the analysis of ventilatory gas exchange. Thirteen healthy adults were submitted to a single maximal oxygen uptake test, being monitored the ventilatory parameters V(O2), V(CO2), and V(E), and the surface EMG (SEMG) of the right vastus lateralis muscle. Each cycle of myoelectric activity was automatically detected and the time series of the mean RMS values of these cycles used for determining the SEMG threshold (SEMG-T) by linear regression. The method was successful to determine SEMG-T for all subjects, with consistent results when compared to respective anaerobic thresholds visually estimated in the ventilatory equivalent V(E)/VO(2).