The development and implementation of the Medicare Diagnosis Related Group (DRG) system has spurred valuable health services research. The focus of this research is often an evaluation of DRGs or another per case type prospective payment system and its capacity to predict resource utilization as well as its potential impact on a wide range of facilities and on patient care. The Wood and Beardmore evaluation of a DRG system for outpatient mental health services provides an additional perspective which enhances the findings of similar studies of psychiatric inpatients. It demonstrates the lack of predictive power and the eventual basic unfairness of implementing a DRG system for psychiatric outpatients. Psychiatric units in general hospitals and psychiatric hospitals are exempt from DRGs. The substantial inaccuracy in psychiatric DRGs has been amply demonstrated and the impact of inappropriate discharge of patients and financial risk to hospitals which treat more severe cases has also been documented. Extending this finding to outpatients is indeed no surprise. The most critical need for research today is for a case mix classification system that accurately measures patient need and cost effective resource consumption. It is important that patients with mental disorders participate in the Medicare prospective payment system, but the DRG system is not an appropriate system. The authors point out that a most critical need is to develop incentives for cost effective use of resources in the continuum from inpatient to outpatient care. A prospective payment system can make a contribution if it can provide the flexibility for clinicians to develop a treatment plan including twenty-four hour hospital care, day treatment and outpatient services, and an adequate up