Restricted immunoreactivity of plasma luteinizing hormone (LH) has been described in some subjects when assayed with certain methods involving antibodies against intact LH. We have compared the performance of the Amerlite LH-30 (A) and Delfia LHSpec (D) assays (which include anti-intact and beta-specific antibodies, respectively) in normal and pathological conditions. As shown previously, results of the two systems were highly correlated with each other and, as we show here, with those of a bioassay. We found eight outliers (results outside the 95% confidence interval of the regression) among 427 samples studied from 121 subjects. Of the outliers, five had Delfia results in a range (< 1 IU/L) that was associated with poor assay precision for that assay, and the ratios of their values by both methods (A:D ratios) were very low. This ratio was affected by endocrine status, e.g., was lower in postmenopausal women than in premenopausal controls, and varied intraindividually within the same menstrual cycle. The restricted immunoreactivity described previously for assays involving anti-intact LH antibodies may, in part, reflect these differences, which, in turn, may reflect the presence of isoforms (e.g., glycoforms) that are differentially recognized by assays that have different antibody configurations.