False-positive serum human chorionic gonadotropin results are estimated to occur in 1 in 10(3) to 1 in 10(4) tests. Most of these false-positive results are due to interference by non-human chorionic gonadotropin substances (especially human luteinizing hormone and anti-animal immunoglobulin antibodies) and the detection of pituitary human chorionic gonadotropin. The false-positive human chorionic gonadotropin measurements are characterized by serum levels that are generally <1000 mIU/mL, the absence of human chorionic gonadotropin in the urine, nonparallelism of the human chorionic gonadotropin levels in serial dilutions of the serum with the human chorionic gonadotropin standard, results that are not consistent with the clinical or operative findings, the absence of a substantial change in the serum levels over time or after therapy, and the finding of a negative result when an alternative type of human chorionic gonadotropin assay is used. An algorithm is presented to aid in the recognition of false-positive human chorionic gonadotropin results in patients.