Background: Retrospective studies show a strong association between child abuse and subsequent psychiatric morbidity. Prospective studies are less common with conflicting data in young adults. We assessed the effect on psychological outcomes at 30-year follow-up of prospective agency notifications compared to retrospective self-reports of childhood maltreatment in the same birth cohort. Methods: We used data on self-reported child abuse from 2425 young adults who completed the Child Trauma Questionnaire (CTQ) at the 30-year follow-up of a prospective birth cohort study commenced in 1981. These were linked to notifications of maltreatment to child protection agencies. The outcomes were DSM-IV diagnoses from the Composite International Diagnostic Interview-Auto version (CIDI). Results: A quarter of the sample (n=600) self-reported maltreatment of any type, 326 (13.4%) rating this as severe. The most common maltreatment type was neglect (n=382), followed by emotional (n=225), sexual (n=198) and physical abuse (n=197). On adjusted analyses, there were differences between agency- and self-reported maltreatment in the psychological effects on 30-year-olds. In the case of depression, and especially post-traumatic disorder, there were significant associations irrespective of reporting sources. In the case of anxiety, there was a strong association with all forms of self-reported maltreatment. However, agency-reported cases were only significantly associated with emotional abuse in the previous 30-days. Limitations: The attrition rate from baseline to follow-up and the possibility of insufficient power to detect statistical significance in agency-reported cases Conclusions: The association between child maltreatment and psychiatric symptoms may vary by diagnosis and reporting source. Each source possibly captures different populations.