Change in mental health status is analyzed as a function of hours of professional treatment, patient attributes, treatment characteristics, and environmental factors for each of seven clusters of conditions typically seen by psychiatrists, psychologists, or social workers. One of two regression strategies was effective in characterizing the relationship between status change and the input variables for each condition cluster. In each model, initial severity was significantly associated with change in mental health status, and it captured most of the unexplained variance. Provider time was predictive of improvement for only some conditions and providers. On average, mental health services appear to make patients better, but improvement is not terribly impressive. Moreover, the marginal benefits of treatment fall off as the number of contact hours increases.