It is a time-consuming task to develop models or empirical equations which would accurately predict fluid properties of polymers as a function of chemical constitution and other experimental variables. Consequently, extensive experimental data have been published in the literature. When such experimental data are available, considerable time can be saved by the use of artificial neural network (ANN) technique based model approaches. Even though these ANNs have been in existence for a long time, their use in predicting fluid properties is still rather limited. In this paper, an attempt has been made to predict the fluid properties of different siloxanes as a function of their molecular configurations with the aid of ANN.