We have described recently the presence of binding sites for human LH (hLH) and human chorionic gonadotrophin (hCG) in microsomal and cytosol fractions prepared from the dimorphic, pathogenic fungus, Candida albicans. We have now compared the properties ofCandida LH/hCG-binding sites with those of the ovine luteal LH receptor. Sheep luteal LH-binding sites were associated with luteal membranes, and little or no binding activity was present in cytosol fractions. In contrast, significant LH/hCG-binding activity was present in Candida cytosol. Moreover, there were marked differences in sensitivity to inhibition by metal ions, association and dissociation rates, and affinity constants between sheep and Candida LH-binding sites. Scatchard plots of 125I-labelled hLH binding to sheep luteal receptors demonstrated a single high-affinity component (association constant (Ka) 0·3 litres/pmol) which was displaceable by hCG. In contrast, Scatchard plots of binding to Candida microsomes and cytosol fractions demonstrated two components, one with high affinity (Ka 0·18 litres/pmol) and low capacity and a second site with lower affinity (Ka 4 litres/nmol) and high capacity, both of which were displaceable by unlabelled hCG. Gel permeation chromatography of cytosol demonstrated two distinct peaks of LHbinding activity with approximate molecular weights of > 1 000 000 and 30 000–50 000. Scatchard plots of 125I-labelled hLH binding to the higher molecular weight peak demonstrated a single, high-affinity LH-binding site (Ka 0·18 litres/pmol), whereas the lower molecular weight fraction contained both high- (Ka 0·17 litres/pmol) and low-affinity (Ka 4 litres/nmol) LH-binding sites. Both partially purified and highly purified hCG and hLH preparations displaced binding of 125I-labelled hLH and hCG to sheep luteal LH receptors at similar concentrations. 125I-Labelled hLH/hCG binding to Candida membranes was also displaceable by low levels (ng) of partially purified hCG preparations, but much higher levels (μg) of highly purified hLH and hCG were required. This paradoxical observation suggested the presence of radiolabelled contaminants, or damaged forms induced during radioiodination of hormone tracers, which can bind more strongly to Candida membranes than unlabelled hCG and hLH but which do not bind to sheep LH receptors. However, no evidence for hLH tracer contaminants with differential binding to Candida and sheep luteal receptors was obtained following gel exclusion chromatography or fractionation on Concanavalin A–Sepharose. (Although three distinct 125I-labelled hLH fractions were resolved on Concanavalin A–Sepharose, presumably reflecting differences in their carbohydrate compositions, all three tracer peaks bound equivalently to both Candida membranes and ovine luteal LH receptors.) Moreover, iodination of highly purified hLH and hCG with 127I failed to generate substances with differential binding to sheep luteal and Candida LH-binding sites. Furthermore, preincubation of 125I-labelled hLH tracer with Candida membranes at different temperatures with or without protease inhibitors did not affect the ability of the hormone tracer to bind to fresh ovine luteal or Candida LH-binding sites. Thus, differences between ovine luteal and Candida LH-binding sites cannot be attributed to differential binding of contaminants present in the hormone tracer preparations, nor to (proteolytic) modification of tracer by Candida membranes during incubation. Journal of Endocrinology (1991) 130, 177–190