Effective discharge planning should serve to reduce length of stay (LOS) and lower readmission rates. A combined prospective/retrospective study was undertaken to assess the effectiveness of anticipated discharge order forms in decreasing LOS for staff pediatric patients. Resident physicians provided advance notice of the planned discharge date to nursing staff and patient family members. Patient selection was random, and the data were collected prospectively. These subjects were matched with patients for whom the order form was not used, and the data was collected retrospectively. Matching variables included: diagnosis, age, gender, race, and method of payment. There were 103 matched pairs. LOS was collapsed into low and high categories. There was a significant difference when comparing subject groups (chi 2 = 8.5; P = 0.005). Whereas only 14% of experimental patients were categorized as high LOS, 32% of control patients were classified as such. Physician prediction of discharge date helped decrease patient LOS.