The combination of air traffic growth and airport capacity limitations has resulted in significant congestion throughout the US National Airspace System, which imposes large costs on the airlines, passengers and society. Absent opportunities for capacity expansion, the mitigation of air traffic congestion requires improvements in (i) the utilization of airport capacity to enhance operating efficiency at the tactical level (i.e., over each day of operations), and/or (ii) the allocation of airport capacity to the airlines to limit over-capacity scheduling at the strategic level (i.e., months in advance of the day of operations). This thesis develops an integrated approach to airport congestion mitigation that jointly optimizes the utilization of airport capacity and the design of airport capacity allocation mechanisms. First, we focus on airport capacity utilization. We formulate an original Dynamic Programming model that optimizes, at the tactical level, the selection of runway configurations and the balancing of arrival and departure service rates to minimize congestion costs, for any given schedule of flights. The model integrates the stochasticity of airport operations into a dynamic decision-making framework. We implement exact and approximate Dynamic Programming algorithms that, in combination, enable the real-time implementation of the model. Results show that optimal policies are path-dependent, i.e., depend on prior decisions and on the stochastic evolution of the system, and that the model can reduce congestion costs, compared to advanced heuristics aimed to replicate typical decisions made in practice and to existing approaches based on deterministic queue dynamics. Second, we integrate the model of airport capacity utilization into a macroscopic queuing model of airport congestion. The resulting model quantifies the relationships between flight schedules, airport capacity and flight delays at the strategic level, while accounting for the way airport capacity utilization procedures can vary tactically to maximize operating efficiency. Results suggest that the model estimates the average departure queue lengths, the variability of departure queue lengths and the average arrival and departure delays at the three major airports in the New York Metroplex relatively well. The application of the model shows that the strong nonlinearities between flight schedules and flight delays observed in practice are captured by the model. Third, we develop an Integrated Capacity Utilization and Scheduling Model (ICUSM) that jointly optimizes scheduling interventions for airport capacity allocation at the strategic level and airport capacity utilization at the tactical level. Scheduling interventions start with a schedule of flights provided by the airlines, and reschedule a selected set of flights to reduce imbalances between demand and capacity, while minimizing interference with airline competitive scheduling. The ICUSM optimizes such interventions, while accounting for the impact of changes in flight schedules on airport operations. It relies on an original modeling architecture that integrates a Stochastic Queuing Model of airport congestion, our Dynamic Programming model of capacity utilization, and an Integer Programming model of scheduling interventions. We develop an iterative solution algorithm that converges in reasonable computational times. Results suggest that substantial delay reductions can be achieved at busy airports through limited changes in airline schedules. It is also shown that the proposed integrated approach to airport congestion mitigation performs significantly better than a typical sequential approach where scheduling and operating decisions are made separately. Last, we build upon the ICUSM to design, optimize and assess non-monetary mechanisms for scheduling interventions that ensure inter-airline equity and enable airline collaboration. Under the proposed mechanism, the airlines would provide their preferred schedules of flights, their network connections, and the relative scheduling flexibility of their flights to a central decision-maker, who may then consider scheduling adjustments to reduce anticipated delays. We develop a lexicographic architecture that optimizes such interventions based on efficiency (i.e., meeting airline scheduling preferences), equity (i.e., balancing scheduling adjustments fairly among the airlines), and on-time performance (i.e., mitigating airport congestion) objectives. Theoretical and computational results suggest that inter-airline equity can be achieved at no, or small, losses in efficiency, and that accounting for airline scheduling preferences can significantly improve the outcome of scheduling interventions.