Objective: While American Academy of Pediatrics guidelines recommend obtaining symptom reports from both parents and teachers when treating children with attention-deficit hyperactivity disorder (ADHD), information from parents is easier to obtain and practitioners may prefer to rely solely on parent report when managing medications. There are, however, few empirical data on the relationship between parent and teacher reports during medication management of ADHD. This study examined the relationship between parent and teacher reports of symptoms of ADHD during a clinical trial. Methods: A study to improve medication management of ADHD was conducted in 24 pediatric practices with 270 children. Children meeting criteria for ADHD were randomized by practice to treatment-as-usual or specialized care groups, with data combined from the groups to examine parent-teacher agreement. Parent and teacher reports on the ADHD Rating Scale-IV were obtained at pretreatment, 4 months, and 12 months follow-up. Results: At each assessment, correlations between parent and teacher ratings were statistically significant, but the magnitudes of the correlations were low, accounting for no more than approximately 17% of the variance between measures. Correlations between change scores on parent and teacher ratings were statistically significant but low for Total and Inattentive scales and not significant for the Hyperactive-Impulsive scale. For agreement on extreme scores, 6 of 9 kappas were statistically significant but all were unacceptably low. Conclusions: Agreement between parent and teacher ratings of symptoms of ADHD is too low for clinicians to rely on parent reports while managing medications. Teacher reports are still needed to ensure optimal management.