Survivors of childhood cancer are at risk for attention problems. The objectives of this study were to assess concordance between parent-, teacher-, and self-report ratings on a measure of attention (Conners Rating Scales, 3rd ed.; Conners, 2008) in adolescent survivors of childhood cancer and to examine associations with a performance-based task. The Conners 3 was completed by 80 survivors of pediatric cancer (39 brain tumor, 41 acute lymphoblastic leukemia; ages 12–17; at least 1 year posttreatment; 51.3% male) as well as their parents and 1 teacher. In addition, survivors completed a continuous performance test. Parents and teachers demonstrated moderate agreement on most subscales; however, agreement was weaker than would be expected based on the normative sample. Agreement between self- and proxy ratings was more variable. The strongest associations for all raters were observed on the Learning Problems subscale. There were significant mean differences between parent and teacher ratings, with parents reporting more problems across subscales. Only self-ratings of Inattention were significantly associated with the continuous performance test (omission errors). Agreement across raters in assessment of attentional functioning in adolescent survivors of childhood cancer is modest. Findings support the need to obtain multiple ratings of behavior, including both proxy- and self-report, when assessing youth with cancer, particularly adolescents.