Adolescent and parent reports of adolescent mental health problems often correlate poorly, and understanding this discrepancy has clinical importance. Yet contextual factors have only been inconsistently explained. At the 14- and 17-year follow-ups of the Western Australian Pregnancy Cohort (Raine) Study, 1,596 parent–child dyads completed the parent-reported Child Behaviour Checklist (CBCL) and the adolescent-rated Youth Self-Report (YSR). Maternal, family, adolescent, and parent factors were examined as potential predictors of discrepancies. When adolescent YSR scores were in the clinical range but parents’ CBCL ratings were not, adolescents were more likely to report alcohol intoxication in the last 6 months, illicit drug use, low school motivation, and depression. When parents reported externalizing behaviour in the clinical range but adolescents did not, the characteristics associated with this were a younger maternal age, receiving social security benefit, stress related to parenting, depression, and poor family functioning. These new results will inform clinical management and research with adolescents who present with behavioural disorders. Statement of contribution What is already known on this subject? We know that adolescent and parent reports of adolescent mental health problems often correlate poorly, but little is known about which contextual factors lead to disagreement. Understanding the factors that influence agreement is clinically relevant for predicting and identifying externalizing behavioural disorders. This is a large-scale study with the ability to assess the impact of numerous psychosocial factors on instrument disagreement. What the present study adds We found that substance use, depression and low school motivation impacted on discrepancy in externalizing behaviour scores for 14-year-old male adolescents and their parents. Parental depression, stress, low family income, and family dysfunction also led to a higher likelihood of discrepancy in scores.