ANN models were utilised in the case of orange (Citrus sinensis L.) cultivation in Adana Province to predict greenhouse gas emissions and total energy input. The share of chemical fertilizers in orange cultivation is high and was calculated as 74.47%. Energy consumption and energy output values in large farms have the highest values with 34 972.38 and 35 305.92 MJ ha^-1, respectively. Furthermore, there is no significant variance in energy use efficiency among the three different sizes of orange farms. In the orange cultivation, non-renewable energy sources (95.88%) have a significantly larger share than renewable energy sources (4.12%). Energy input in direct energy, indirect energy, renewable energy, and non-renewable energy sources configurations was calculated as 6 819.99, 27 185.07, 1401.98, and 32 603.08 MJ ha^-1. The greenhouse gas analysis showed greenhouse gas emissions equivalent to 759.58 kgCO_2eq ha^-1. 56.84% of greenhouse gas emissions come from chemical fertilizers. The best Artificial neural network model training data used for orange production and greenhouse gas emissions has root mean square error values of 0.141 and 0.063, respectively, while the mean absolute percentage error values are 0.005 and 0.004, respectively. Due to the ability of Artificial neural networks to predict results, it can be effectively used in growing oranges and other plant crops.