With the complete construction of the All-Roundly Well-off Society in China, residents have higher and higher requirements for the quality of life, especially for food safety. However, food quality and safety accidents such as “deteriorated meat”, “adulterated meat”, “added meat” and “water-injected meat” frequently occurring to threaten the life safety of Chinese residents seriously and hinder healthy development of the market. The quality test method of red meat is a physical and chemical experiment that seriously damages the samples and is only applicable to the spot check of the market supervision department. Hyperspectral technology is a kind of in-situ non-destructive, high-throughput and, fast intelligent detection technology which provides effective technology for solving the low operational feasibility of traditional detection methods. It greatly promotes the development and improvement of the quality and safety supervision system of red meat in China. This paper aims to review the research progress of hyperspectral technology in non-destructive detection of red meat quality. Firstly, the advantages and disadvantages of the red meat quality model based on the Hyperspectral technique are summarized. Its advantage is high resolution and a combination of image and spectrum, which will provide better data for the model. Then, the key algorithms in the model are analyzed: (1) Due to regions of interest obtained manually, automatic separation of regions of interest will be one of the focus of research; (2) The spectral preprocessing algorithm is mainly selected by observing the spectral signal or extrapolating by model, so there is no standard general preprocessing algorithm; (3) The combination of spectrum and image features can more comprehensively describe the quality of red meat and provide a batter basis for modeling; (4) The linear model is more mature and stable, but the research potential of nonlinear model is better for the complex environmental factors in red meat quality detection. Finally, the future development direction and research focus of hyperspectral technology in red meat quality prospect. Finally, the key research direction of hyperspectral non-destructive detection for red meat quality is concluded as improving algorithm automation, making full use of spectrum information and strengthening the application of the nonlinear model based on the summary of the research results in recent years.