After a brief review of the known techniques for the identification of crossed immunoelectrophoresis peaks, the authors claim that the technique of Platt et al (1973), Clin. Chim. Acta 46, 419-429) is the most effective. This consists of blocking a protein in the first run, thereby causing the disappearance of the corresponding peak in the second dimension. The advantages of this technique with respect to the others are discussed.