Over the last three decades, fibre has been commonly used in reinforced concrete (RC) columns to enhance their strength and ductility. In the available studies on Steel Fibre Reinforced Concrete (SFRC) columns, axial compressive strength of the columns is expressed as functions of unconfined concrete compressive strength and the lateral confinement pressure. In this study, available experimental investigation results of the axial compressive strength of SFRC square columns have been compiled. This study also proposes axial compressive strength models of the SFRC square columns using artificial neural network (ANN). In the proposed models, axial compressive strength of SFRC columns has been expressed as functions of unconfined concrete compressive strength (fc'), confinement pressure due to lateral reinforcement (fle) and confinement pressure due to the action of steel fibres (flb). The results of the proposed ANN strength models match very well with the available experimental results.