Immune and non-immune rubella sera were examined in the haemagglutination inhibition (HAI) test to evaluate the removal of lipoprotein non-specific inhibitors (NSI) of haemagglutination after kaolin and manganous chloride/heparin treatments. The sera were titrated fresh or after lipoprotein deterioration brought about by ageing the samples at 4 degrees C for various periods of time and by freezing at -20 degrees C with subsequent thawing. Deterioration was seen as altered electrophoretic mobility while the lipoproteins in treated sera were detected by indicator arboviruses whose haemagglutination is known to be strongly inhibited by the native macromolecules. After both treatments, notably manganous chloride/heparin, residual NSI activity was detected in deteriorated samples, particularly with group B arboviruses such as Japanese encephalitis and west Nile viruses but generally less so or not at all with the group A arboviruses employed. Absolutely fresh sera are considered highly desirable for rubella virus HAI assay, and it is suggested that the efficiency of lipoprotein NSI removal regardless of treatment protocol could be monitored in parallel HAI tests using carefully chosen indicator arboviruses. This could be done in conjunction with density gradient centrifugation of doubtful sera should ultracentrifuging facilities be available. The suitability of the monitoring procedure would be dependent to some extent on whether certain arboviruses are known to be endemic in a particular area.