Non-scheduled air transportation, characterized by its dynamic and non-scheduled passenger flows, presents unique challenges for efficient resource allocation and operational planning. This paper investigates the application of stochastic forecasting models to predict passenger flows in non-scheduled air transportation, accounting for inherent uncertainties in demand patterns, market dynamics, and external factors such as weather and regulatory changes. We propose a comprehensive framework that integrates stochastic processes, time series analysis, and probabilistic simulation techniques to model and forecast passenger flows with quantified uncertainty. By leveraging historical data and incorporating real-time information streams, our approach enables operators to anticipate demand variations, optimize fleet utilization, and improve service reliability. Through case studies and empirical validation, we demonstrate the effectiveness of stochastic forecasting models in capturing the complex dynamics of passenger flows in non-scheduled air transportation, providing valuable insights for operators to enhance operational efficiency and adaptability in a rapidly changing aviation landscape.