In a previous study guinea pigs inbred for their ability to develop respiratory anaphylaxis to experimental antigens have been used for comparison of different forms of immunotherapy (IT). Passive, active and combined (immune complexes prepared from antigen and specific IgG) IT was compared with placebo. In the present study methods were evaluated for determination of the allergen-specific IgE and IgG. IgE was determined by the passive cutaneous anaphylactic test (PCA) and the variability of this test on different strains of the recipient guinea pig was investigated. The same strain as used for the IT study was found to produce the most potent response. Radioimmunometric assays (RIA) were developed and validated for determination of specific IgG1 and IgG2. The IgE and IgG immune response in animals from the IT study were then evaluated by means of PCA and RIA. Animals from all four treatment groups were sensitized during the first part of the IT study, and responded with a marked IgE synthesis which later stabilized on a more moderate level. In spite of notably reduced symptoms in groups treated with active and combined IT, no difference in the IgE level was found between the four groups. In contrast to IgE, mean group titers of IgG1 and IgG2 in the groups receiving active or combined IT rose drastically during the first part of therapy and closely paralleled the clinical response during the rest of the study period. However, in the individual animals, no correlations were found between immune response and clinical symptoms. Thus, the strong IgG response during immunotherapy may not be causally related to the outcome of treatment.