Discrimination of structurally similar analytes by immunoassay is limited by antibody cross reactivity. Using a plurality of cross-reacting antibody species allows increased selectivity by application of pattern recognition methods. We present a detailed characterization of an array of monoclonal antibodies which allows analytical modeling of the performance of an antibody array in a multi-analyte system. Such well defined antibody arrays give the possibility for the systematical optimization for immunoassay applications. Affinity characterization is carried out in a simple test format: After equilibrium binding of antibody and analyte, unoccupied antibody is quantified by an optical transducer. The test result reflects directly the respective affinity constants for different analytes. A set of three monoclonal antibodies was characterized with respect to their affinity to five different s-triazines which play an important role in water contamination. The affinities were compared with results obtained by direct enzyme immunoassay. The analytical performance of the antibody array was modeled by using the affinity constants determined from the calibration curve.