The research objective of this thesis is to develop an innovative airline fleet planning concept that is capable to consider the long-term stochastic nature of air travel demand while generating meaningful results in reasonable computation times. The proposed methodology aims to identify robust fleets, in terms of profit generating capability across a long-term planning horizon under stochastic demand, through the adoption of a portfolio of fleets (each of different size or composition) and a three-step modeling framework. The three models involve the simulation and sampling of stochastic demand using the mean reverting Ornstein-Uhlenbeck process, iteration over an optimization model that optimally allocates each fleet from the portfolio given the demand sample values, and a scenario generation model that generates scenarios across the planning horizon. A case study is presented and serves as proof of concept.