A central question in the study of executive politics concerns whether presidents behave in an idiopathic, opportunistic fashion reflecting individual differences across administrations, or whether presidents behave as institutional actors who are constrained by external forces that occur across administrations. We argue that development of the institutional presidency is an important consideration in understanding presidential behavior. We maintain that presidents will behave consistent with the opportunistic model during the developmental phase of the institutional presidency, while they will act compatibly with the constraint model once the institution has fully matured. Using data on executive orders for the 1939–96 sample period, both event count regression techniques and model selection criteria are employed to test our theory. The statistical evidence supports our claim that each theory has empirical merit, depending upon the stage of the institutional presidency under investigation.