This study aimed to evaluate the efficiency of energy consumption and economic analysis of different watermelon cultivation systems in Fars Province of Iran. Watermelon production systems were classified into five systems, namely, custom tillage (group 1), conservation tillage (group 2), traditional planting (group3), semi mechanized planting (group 4), and mechanized planting (group 5). Data were collected from 317 watermelon producers from different parts of the province through face to face interviews. Multi-Layer Perceptron artificial neural networks were used to model the energy flows of watermelon production. The results showed that the greatest energy consumption belonged to mechanized planting system with the value of 81317.72 MJha-1 and with the productivity of 0.61 kgha-1 and energy use efficiency of 1.17. Clustering function with three inputs (human resources, machines and diesel fuel) showed that the difference between groups 2 and 4 is more than the other groups. The least energy consumption belonged to the conservative agriculture as78163.86 MJha-1 and the energy productivity and energy use efficiency about 0.64 kgha-1 and 1.22, respectively. The results of energy modeling showed that an ANN model with 9-10-1 structure was determined to be optimal for energy flow modeling of this system. Generally, it was concluded that the artificial neural network models can be applicable to prognosticate the energy flows of watermelon production. From an economic point of view, the least net profit belonged to traditional planting with the value of 2618.14$, and the most net return belonged to mechanized planting with the value of 2752.88$/ha.