Uncertainties in chemical process performance behaviour continue to cause considerable concern to engineers and other stakeholders. The traditional deterministic uncertainty modelling methods lead to excess overdesign, which is expensive, and have also been shown to give limited insight into the behaviour of complex chemical engineering systems. The present work develops a new framework, termed “Stochastic Process Performance Modelling Framework (SPPMF)”, which combines traditional deterministic process simulation, response surface modelling techniques and advanced structural reliability analysis methods to facilitate efficient performance modelling and optimisation of chemical process systems under uncertainties. Cross application of structural reliability principles to chemical processes presents some challenges; however, means of addressing such issues are proposed and discussed in this thesis. For instance, to facilitate Process Reliability Analysis (PRA), stochastic constraints have been added to the conventional process optimisation formulation. Both first order reliability method and Monte Carlo simulation are then applied to gain a wide range of performance measures. In addition, to allow for automated response surface generation, an interface for linking process simulators and a new stochastic module has been developed; making it possible to obtain samples in the order of thousands, typically in minutes. A number of Structural Reliability Analysis (SRA) concepts have been re-defined to reflect the unique characteristics of chemical processes. For example, while SRA is mainly concerned with the effects of random forces and mechanical properties on structural performance, PRA is focused on random process conditions (e.g. changes in pH, reaction rates, etc) and their effects on both product quantity and quality. Finally, SPPMF has been successfully applied to model stochastic properties of a range of typical process systems. The results show that the new framework can be efficiently implemented in process engineering with significant benefits over the traditional methods. Limitations of SPPMF and directions for future work are also highlighted. This thesis contains commercially confidential information which should not be divulged to any third party without the written consent of the author.