The research was conducted for a representative 50 ha farm in the Onderberg region in Mpumalanga province, where farmers use a combination of centre-pivot, drip, and dragline systems of different sizes to grow sugarcane. The main intention was to establish a multi-period linear programming model capable of economically evaluating a farm's expansion decision-making process for farmers faced with investment decisions in alternative irrigation systems, taking into account the available initial capital of the farm. A linear programming (LP) model was used to assign a mainline for a total of twelve irrigation system combinations based on the assumption that the farmer wishes to start with a 30 ha centre-pivot investment. The Generalized Algebraic Modelling System (GAMS) was used to formulate the farm growth model as mixed integer dynamic linear programming (MIDLP) for a 15-year planning horizon. Based on the results, farmers are initially forced to invest in lower-cost irrigation systems when they lack capital to start a farm business due to the time value of money. They only consider lowering operating costs by investing in capital intensive irrigation systems when they have more own capital or borrowing capacity.