One hundred randomly chosen sera from blood donors from North London were assayed for antibodies to mumps virus by plaque reduction and microtitre neutralisation assay, haemagglutination inhibition and in-house ELISA. The assay reproducibility was determined, and there was reasonable agreement between antibody levels measured by the two neutralising methods. Neutralising antibody levels measured by either method were low but the strain used in the assay had a large effect on the antibody titres observed. Titres measured by neutralisation assay, HI assay and ELISA did not correlate well. Assessment of immunity to mumps virus remains problematical.