Male athletes may administer human chorionic gonadotropin (hCG) to enhance performance. Despite imposing a ban on hCG doping, the International Olympic Committee has no recommended procedure to confirm the presence of hCG. Our aim was to establish a limit for hCG in urine above which a sample may be considered positive. We measured hCG concentrations in urine samples from 1400 men with the Serono MAIAclone IRMA. Statistical evaluation of the results gave a "far outside" value [75th percentile + (3 x interquartile range)] of 5.0 IU/L; greater values are extremely unusual. Immunoreactive material in urine samples from a reference group of 120 noncompeting individuals was concentrated about sevenfold by centrifugal ultrafiltration; all these concentrated samples gave assay values < 5.0 IU/L. To ensure with the greatest possible degree of certainty that no false-positive result is reported, we propose a decision limit of 10 IU/L in nonconcentrated ultrafiltered samples. We also prepared quality-control material to contain hCG at 10 IU/L. This material provides a decision limit above which, after confirmatory procedures, a sample should be considered positive. This should help establish comparability of results among laboratories testing for sports drugs.