Concrete Filled Steel Tubes (CFST) are a composite building material widely used in modern structural design, providing benefits over conventional structural members. As a composite material, the design methods for calculating the strength of such columns and beams has not yet received the level of development that conventional single materials have. This project seeks to develop a design tool to aid the standard design process and achieve more accurate strength predictions. Using machine learning, specifically Artificial Neural Networks (ANN), several prediction models were developed to analyse the physical and mechanical properties of circular CFSTs to determine the axial strength of the column. Cross-terms; model parameters developed from the physical and mechanical properties were used in developing the separate ANN models. These cross-terms were to analyse the relationship between the different physical properties. The developed models showed improved accuracy in their predictions, over the current conventional design methods.