The 125I C1q deviation test and the modified 125I C1q PEG precipitation test were compared in their ability to detect soluble immune complexes in serum using a model system of HSA-rabbit-anti-HSA, and were then applied to sera collected from patients with malignant and non-malignant conditions. Despite close agreement in the model system, the two tests gave divergent results for the presence of C1q binding substances in individual serum samples collected from patients. The inherent complexities of interpreting C1q binding in serum, in terms of the presence of soluble immune complexes, makes it questionable whether either test can be relied upon to provide a means of identifying these complexes in the sera from patients with malignant disease.