The study objective was to determine the clinical value of positive antinuclear antibody (ANA) and ANA profile tests in children with autoimmune disorders. A retrospective chart review was carried out of all patients under 18 years of age with a positive ANA test (HEp-2 cell substrate, titre > or =1:40) and ANA profile (ELISA) referred to the paediatric rheumatology service at the authors' institution between 1992 and 1996. Of 245 children with a positive ANA test, 134 (55%) had an autoimmune disease, including juvenile rheumatoid arthritis (n = 49), systemic lupus erythematosus (SLE) (n = 40) and others (n = 45). The remaining 111 patients did not have identifiable autoimmune diseases. Patients with autoimmune disorders had significantly higher ANA titres of > or = 1:160 (chi2 = 16, P<0.0001). In addition, of the 245 patients with a positive ANA test, 86 had an ANA profile performed; this was positive in 32 and negative in 54. All 32 patients with a positive ANA profile (100%) had an autoimmune disorder, compared to 22 (41%) of 54 with a negative ANA profile who had autoimmune disorders. Of 22 SLE patients with a positive ANA profile, 16 (73%) had positive anti-dsDNA and 15 (68%) had positive anti-Sm and positive anti-RNP. A positive ANA profile correlated strongly with an ANA titre > or = 1:640 (chi2 = 5.7 , P<0.02). The study demonstrated that only 55% of children with a positive ANA test had a definitive diagnosis of autoimmune disorder. These children tend to have higher ANA titres of > or =1:160. However, a positive ANA profile was strongly correlated with an ANA titre > or =1:640 and highly indicative of an autoimmune disorder (100%). We suggest that in order to reduce cost, an ANA profile should not be performed on all patients with positive ANA, but reserved for those with an ANA titre of > or =1:640 and/or those with a high clinical index of suspicion for autoimmune disorder, especially SLE.