Insight into the consequences of the in vivo interaction between antigen and antibody depends on the analysis of systems in which relatively simple chains of reactions can be studied. One such system is passive cutaneous anaphylaxis (PCA), an immediate type of skin reaction (1, 2), which is initiated by a combination between antigen and cell-fixed antibody in the skin and characterized by increased permeability of the walls of minute blood vessels in the skin. The first of the events leading to this increase in permeability is the attachment of antibody to cells in the intradermally injected site of the skin (3-7). The next event is the attachment of antigen to the fixed antibody. After this a series of further reactions culminates in increased permeability of blood vessels. The question arises as to what these further reactions arc, and, in particular, whether complement is involved in them. So far, most attempts to elucidate this question were based on experiments with animals in which the circulating level of complement was reduced experimentally. Such reduction can be caused by systemic injection of antigen-antibody precipitates before the injection of the reactants, leading to PCA (8-10). Another approach has been based on chemical modifications of antigen which result in reduced complement fixation by the complex of antigen and antibody. On the basis of such experiments it was concluded that complement is involved in PCA of the rat (8, 9), but not of mice (11).The role of complement in PCA of the guinea pig was examined by a third type of analysis: it is possible to separate antibody into a fraction which fixes complement and another which does not. Experiments with such preparations are consistent with the view that complement is not involved in PCA of the guinea pig (10).There are other cutaneous immunological reactions in which a role of complement has been postulated; one of these is the reverse passive Arthus reaction (RPAR), so far studied in the rat (12), rabbit (13,14), guinea pig (7,10,15), and mouse (16). The R_PAR has been differentiated from PCA by its histopathology, is caused by aggregates of antigen and antibody, and does not involve the fixation of antibody to cells (17,18). It is thus characterized by the fact that it can be elicited by antigen, administered simultaneously with the intradermaliy injected antibody. Unlike PCA, R_PAR does not depend on the fixation of antibody on cellular sites. In the context of our present