Abstract: With accelerated development of edible fungi industrialization and intelligent cultivation, growth models have become key tools for environmental control, yield prediction, and quality management．Focusing on biological chain of substrate-mycelium-fruiting body, three technical routes and their application progress were systematically reviewed, including empirical models, mechanistic models, and intelligent models．Results indicated that empirical models were easy to construct and provide rapid predictions, making them suitable for process optimization and online release．However, they lacked sufficient physiological explanation and cross-scenario generalization．Mechanism models describe carbon-nitrogen balance, reserve mobilization, maintenance consumption, and threshold triggering through a three-library concept, offering explanations for flush cycles and yield formation, but of parameter determination and calibration cost was high．Intelligent models relied on time-series forecasting, object detection, and closed-loop control to identify growth stages identification, quality classification, and robotic harvesting, with outstanding accuracy and real-time performance．However, they were constrained by data scale, interpretability, and transferability．Overall, research was shifting from an experience-driven approach toward a mechanism and data fusion gray-box paradigm, while deeply integrating with internet of things（IoT）, edge computing, and model predictive control．Future efforts should focus on developing standardized datasets and evaluation systems accross varieties and scenarios, enhancing interpretability constraints and uncertainty reporting, advancing digital twins and multimodal perception, and promoting large-scale application of growth models in standardized, replicable, and low-energy production．