Current methods to establish the duration of toxoplasma infection in pregnant women and for the diagnosis of toxoplasmosis in the neonate or HIV infected patient have significant limitations. We assessed the precision of a commercial ELISA for the detection of toxoplasma specific IgG and adapted the assay to measure avidity using an elution agent washing step. The sensitivity and specificity of the ELISA were 100 and 75% respectively and optimal measurement of avidity was achieved using 6 M urea as the elution agent. Toxoplasma lymphadenopathy of less than 3 months duration was associated with low avidity specific IgG but some discordant findings were recorded. Serial measurement of IgG avidity assisted the distinction between actively produced antibody in infants with congenital toxoplasmosis and passively acquired antibody of maternal origin in uninfected babies. There was no significant difference between avidity levels in HIV infected patients with or without cerebral toxoplasmosis.