We used a "sandwich"-type immunoenzymometric assay (IEMA) and a radioimmunoassay (RIA) to measure antibody against the human nicotinic acetylcholine receptor in serum from individuals with myasthenia gravis, with markedly different results for certain specimens, as measured by the two techniques. In some cases, antibody concentrations were high by RIA but low by IEMA; in others, the reverse was found. Such differences persisted through 30 months after thymectomy. An investigation of potential causes of this disparity suggests that high IEMA measurements reflect specific anti-receptor antibody and are not artifactual. The IEMA is recommended as an adjunct to the RIA because some patients with myasthenia gravis who have low concentrations of anti-receptor antibodies as measured by RIA have significantly above-normal concentrations of anti-receptor antibodies as measured by IEMA.