Sixty-four subjects' serum samples, positive or equivocal by rubella IgM assays and containing rubella IgG, were examined for the avidity of rubella IgG. Four of the sera originated from rubella reinfections; others had false-positive IgM results due to interference by parvovirus infection or by other mechanisms; and the remaining were sera from the acute phase or convalescence of primary rubella. A novel IgG avidity test, avidity-ELISA, and a semiquantitative haemolysis typing assay were used. According to the avidity-ELISA, 29 subjects had recent primary rubella (low IgG avidity), and another 29 had previous rubella immunity (high IgG avidity), whereas 6 serum samples gave borderline avidity values. Comparison of these results with pre-existing clinical records and laboratory data showed that all samples with low IgG avidity were obtained during or shortly after acute primary rubella. All sera with high IgG avidity originated from the previously immune subjects; the rubella reinfections were confined within this group. Five of the six sera with borderline avidity values were obtained within 2 months from primary rubella. In conclusion, the measurement of IgG avidity is a powerful tool for the distinction of acute or recent primary rubella from pre-existing rubella immunity, including rubella reinfections.