This report describes the results of the first part of the collaborative study organized by a working group sponsored by the Community Bureau of Reference of the European Community Commission. The whole study was designed to understand the causes of discrepancy among LH immunoassay methods. In the parent work, we studied the characteristics of 55 monoclonal antibodies to LH which allowed us to establish a detailed map of the antigenic surface of the hormone. In the present report we used this information to interpret the discrepancy in LH concentrations assayed with 12 different methods in 300 sera from subjects with various clinical conditions. The 55 monoclonal antibodies provided by 11 commercial companies were tested in various experiments: the apparent affinity of the antibodies was, generally but not always, higher for LH presented by a second antibody coated to plastic than for LH directly coated to plastic; 26% of the antibodies recognized the alpha subunit, 26% the beta subunit and 48% reacted only with the holomolecule (anti-alpha beta); only 28% of the antibodies were strictly specific for LH. Criss-cross experiments allowed us to distinguish 13 antigenic regions on the surface of LH: 6 were located on the alpha subunit, 3 on the beta subunit and 4 on the holomolecule. The monoclonal antibodies to the alpha beta regions further separated into 12 clusters of reactivity. Accordingly, LH appeared to exhibit at least 21 epitopes. Comparison of the immunoreactivity of various LH preparations indicated that highly purified pituitary LH and immunoaffinity purified urinary LH reacted similarly with the monoclonal antibodies and strongly differed from crude urinary LH. These data indicated that the immunoreactivity of an LH preparation depends mainly upon the degree of purification and not that much upon the origin of the preparation. The epitope specificity of the monoclonal antibodies used in 11 commercially available LH assay kits was also determined: 10 kits used at least one anti-alpha beta monoclonal antibody associated with an anti-beta monoclonal antibody in 7 cases or another anti-alpha beta monoclonal antibody in 3 cases; one kit used an anti-alpha monoclonal antibody associated with an anti-beta monoclonal antibody. None of the kits were strictly identical with regards to the epitope specificity of the monoclonal antibodies used.