Abstract: Accurate monitoring of endurance training load requires translating physiological intensity boundaries established in the laboratory into executable, trackable external-load parameters in field-based training. Functional threshold power (FTP) is associated with maximal lactate steady state (MLSS) and other metabolic thresholds; however, their consistency is affected by the testing protocol, anaerobic work capacity, and training status, and the corresponding values should not be used interchangeably. Based on the research evidences from cycling, this study reviewed the theoretical development and physiological basis of FTP, distinguished it from MLSS, lactate threshold, ventilatory threshold, and critical power, and then summarized assessment approaches including prolonged-duration testing, the 20-min maximal sustainable power test（FTP₂₀）, and historical power-data modeling. Particular attention was given to the anaerobic compensation error associated with the conventional 95% conversion factor and to individualized calibration. In this study, we further explained FTP-based intensity zoning and periodized load regulation, proposed a dynamic calibration pathway integrating power output, muscle oxygen saturation, and heart rate variability, and discussed its practical application through observational cases involving cycling-specific training and land-based auxiliary training for speed skating. Current evidence indicated that the most established applications of FTP remain concentrated in cycling, whereas other endurance sports require independent validation within their sport-specific threshold systems. The principal value of FTP lies in bridging laboratory physiological assessment and field-based power training by providing a practical reference for individualized intensity zoning, periodic reassessment, and periodized load adjustment, rather than replacing physiological gold standards such as MLSS. Future research should employ large-sample controlled designs, cross-sport calibration, and sport-specific power modeling to clarify its validity boundaries and contribution to training adaptation.