The performance status of the child with cancer is an important outcome consideration in pediatric oncology research and practice. However, no single measure for children has been available. This is a report of the development and standardization of such a scale. The play-performance scale for children is a parent-rated instrument which records usual play activity as the index of performance. Performance status ratings were obtained on three groups of children: patients (n = 98), patients' siblings (n = 29), and an independent sample of hospital employees' children (n = 40). Children with all types and stages of childhood malignant neoplasms were represented. Test results established the parent as a competent, reliable rater and demonstrated the validity of the scale. Interrater reliability was examined using correlational statistics and percentage agreement. Agreement between parents was good, and there were no systematic rater biases. In addition, parents' ratings significantly discriminated differences in levels of functioning (mean score, patients 75.4 versus siblings 97.4). Correlational and analysis of variance (ANOVA) procedures demonstrated that the play-performance scale was significantly related to the global performance measures of experienced clinicians and was sensitive to change. Inpatients received a mean score of 42.3, outpatients 90.7, and normals 98.2. These findings indicate that the scale is both feasible and effective. It is concise, can be administered repeatedly even to extremely ill patients, and uses parents as observer reporters. The play-performance scale for children provides quantifiable, reproducible, and meaningful data, which is necessary for effective monitoring and management of the child with cancer.