In recent years, the outbreaks of foodborne diseases caused by pathogenic bacteria have made considerable economic losses and shown a threat to public health. The key to prevent and control these diseases is fast screening of pathogenic bacteria, which is usually performed with three procedures: sample collection, bacteria separation and bacteria detection. For sample collection, the national standard methods are often employed. For bacteria detection, currently available methods such as Polymerase Chain Reaction and Enzyme Linked Immuno-Sorbent Assay are often used. For bacteria separation, traditional methods such as filtration and centrifugation are not capable to specifically separate the target bacteria. However, food samples are very complicated and require efficient pretreatment for bacteria separation and concentration to achieve accurate and reliable results. The conventional immune magnetic separation method can be used to specifically separate the bacteria, but it still cannot meet the requirements for food sample pretreatment due to very low concentration of target bacteria in food. Therefore, this study developed an automatic and efficient immuno-separator of foodborne bacteria based on magnetophoresis and magnetic mixing, and E. coli O157:H7 was used as research model. A magnetic mixer was applied to facilitate the immunoreaction between the immune magnetic nanoparticles and the target bacteria cells, and a magnetophoretic separation tubing was utilized for magnetophoretic separation of the magnetic bacteria. Under the optimal mixing time of 20 min and the optimal flow rate of 50 µL/min, the separation efficiency of E. coli O157:H7 could be more than 90%, showing that the developed immuno-separator is promising to be applied for efficient separation of foodborne bacteria and can be easily extended for separation of other biological targets by using their specific antibodies.