Empirical evidence consistently finds little or no impact on costs related to the introduction of parity. Busch and colleagues in this issue add to the analysis of parity (5). A difference-in-difference analysis of utilization and cost data from the evaluation of the Federal Employee Health Plan (FEHP) parity policy examined the effects of diagnosis before and after the implementation of parity relative to employed individuals who were not federal employees. “Difference in difference” means that for both the federal employees and the control nonfederal employees, data from an earlier period when both groups lacked parity were compared with data from a later period when the federal plan had parity and the control plans did not. The differences between the two periods for each group were then compared for the final analysis, a difference between groups in the difference between the two time periods. Three mental health diagnostic groups that typically differ in severity were extracted from the data set—individuals with bipolar disorder, the most severely ill diagnostic group (federal, N=2,557; comparison, N=1,177), individuals with major depression (N=10,412 and N=5,245, respectively), and individuals with adjustment disorder, the least severely ill diagnostic group (N=6,125 and N=4,099).