To provide a high through-put method for protein identification by which necessity of cutting out of spots is eliminated, a time required for protein identification is shortened and a high-speed on-line operation is continuously carried out and an apparatus therefor.

The method for protein identification comprises separating and purifying various kinds of proteins by a separation and purification method not depending upon a gel and subjecting a separated and purified substance to an amino acid sequence analysis to identify the proteins. The separation and purification process comprises at least two stages of separation and purification processes and the first-stage separation and purification process is carried out by free-flow electrophoresis method.