Informant-related determinants of item attenuation, that is, the drop-off in symptom endorsement rates at retest, were examined in an enriched community subsample of 245 parent-child pairs drawn from the National Institute of Mental Health Methods for Epidemiology of Child and Adolescent Mental Disorders Study. Youngsters and their parents were interviewed with the Diagnostic Interview Schedule for Children (Version 2.3; DISC-2.3) on two occasions with a mean test-retest interval of 12 days. Item attenuation rates were high for both informants, with adults failing to confirm 42% and children 58% of baseline responses at retest. Stepwise regressions revealed that item attenuation at DISC-P retest was higher for adult informants who were younger, and who reported on older and less impaired children. On the DISC-C, attenuation was higher for children who were less impaired, rated as doing worse in school, and who had a longer test-retest interval. These results are broadly consistent with past studies examining the determinants of attenuation and test-retest reliability and have implications for the design and use of structured diagnostic instruments.