• parents;
  • school teachers;
  • risk;
  • prognosis;
  • child;
  • psychopathology

Objective:  To investigate if parent–teacher discrepancies in reports of behavioral/emotional problems in children predict poor outcome.

Method:  A total of 1154 4- to 12-year-old children from the general population were followed up. At the first assessment, parent and teacher ratings were obtained with the Child Behavior Checklist (CBCL) and Teacher's Report Form (TRF). Fourteen years later, DSM-IV diagnoses were assessed, and ratings of self-reported and parent-rated behavioral and emotional problems were obtained.

Results:  CBCL and TRF scores predicted most of the outcomes, but in general, discrepancies between CBCL and TRF scores did not. There were some exceptions. For instance, higher parental vs. teacher ratings of aggressive behaviors increased the risk of suicide attempts/self-mutilation.

Conclusion:  Risk factors for self-mutilating behaviors may be supplemented with parent-reported aggressive behaviors that are not observed by the teachers. In general, whereas CBCL and TRF scale scores were useful predictors of outcome, parent–teacher discrepancies were not.