Abstract: Intellectualization is the requirement and inevitable trend of mineral processing development. The traditional detection technology of mineral processing parameters has strong subjectivity, large error and poor timeliness. Online detection of mineral processing parameters is the basis for achieving stable operation, optimal control and intelligent management of mineral processing plants. The mineral processing process mainly includes four steps: crushing−screening, grinding−classification, separation and dewatering. The separation includes gravity separation, magnetic separation, flotation and other methods. The online detection technology of key process parameters such as grade of ore, particle size, solids concentration and flow of pulp, flotation froth properties and moisture content of concentrates were summarized. The potential problems and developmental prospects of online detection technology of key parameters in mineral processing were pointed out. The existing online detection technology is mostly quantitative detection of single process parameters. The single parameters detection technology is difficult to adapt to the complex mineral processing process and cannot meet the requirements of automation and intelligent development. The lack of advanced online detection technology for mineral processing process parameters restricted the construction of intelligent mineral processing plants. With the development of automation and intelligent technology, online detection technology of multi parameters is imminent. In the future, researchers need to strengthen the research in these aspects: (1) the content of trace key metal elements in ore and the chemical properties of pulp during grinding; (2) the size distribution of flotation bubble, gas dissipation rate, flow field distribution, and residence time distribution of pulp; (3) the rapid and accurate detection technology of the types and concentrations of residual reagents and harmful pollutants in circulating water system. Researchers could develop online detection technology of multi parameters combined advanced ultrasonic technology, optoelectronic technology, infrared technology, terahertz technology, optical fiber technology and nuclear technology. The accuracy of artificial intelligence algorithm and calculation model can to be improved through the machine learning and deep learning technology. The key of artificial intelligence simulation and prediction in mineral processing is to collect and form large data sets with the help of multi parameter detection instruments and intelligent control platforms. The development of multi parameters online detection technology is important to the construction of intelligent and unmanned mineral processing plants. This review will help much people understand the current situation and development trend of mineral processing parameters online detection technology.