The concentration of furanic compounds in transformer's oil can be an effective measurement towards assessing the aging state of oil impregnated paper in the transformer. The rate of change of the concentration of furan content is vital for assessing the rate of deterioration of cellulose insulation and its severity. This promotes furan content as effective parameter in transformer oil for transformer condition assessment and accordingly asset management. In this paper the correlation between oil parameters and furan content is studied using artificial neural networks (ANN). A neural network is used for predicting the furan content based on different combinations of input parameters that are known to be correlated to cellulose paper degradation of the transformer. These input parameters are carbon monoxide (CO), carbon dioxide (CO2), water content, acidity, and break down voltage (BDV). Results on real data of forty transformers show that the proposed model is capable of predicting the furan content with an average accuracy of 90%. Consequently, this proposed model improves the efficiency of oil chemical tests and dissolved gas analysis (DGA) and their abilities to assess the condition of transformer solid insulation.