It is proposed that the multiple-component neural network model be extended to make use of multiple-component multiple-factor analysis. First, a neural network model and a linear multiple regression model are developed and compared using multiple-component analysis to demonstrate the better modelling technique for pork farm odour. The odour samples were collected using a vacuum box fitted with a pump to draw air into 10 L Tedlar bags. The odour dilution threshold was determined within 48 h of sampling using trained human assessors and a dynamic olfactometer. Approximately 20% of the samples, or 26 data points, were randomly selected for model testing and the remaining 80% of the samples, or 105 data points, were used for model development. The neural network model of the pork farm odour yielded more accurate and precise odour intensity predictions than the linear multiple regression models, indicating that neural networks are the better modelling technique for this application. Subsequently, a multiple-component multiple-factor neural network model was developed and compared with the multiple-component neural network. The multiple-component multiple-factor neural network model generated performance gains, indicating that this approach is relevant to modelling pork farm odour. The multiplecomponent neural network model provided better performance than the corresponding linear multiple regression model. This demonstrated that multiple component farm odour models benefit from the use of contemporary intelligent modelling techniques, specifically neural networks.