Abstract <p>The influence of plastic viscosity, a vital rheological property of drilling fluids, is particularly important in determining the cleaning of the borehole and the efficiency of the drilling processes. To prevent operational difficulties, the plastic viscosity must be measured accurately and in a timely manner. Unfortunately, conventional laboratory techniques employed for its assessment are challenging to complete as the drilling progresses. To meet this need, this research introduces a novel artificial neural network (ANN) model that estimates plastic viscosity based on six commonly measured mud properties: March funnel viscosity, mud density, solid content, water content, oil content, and salinity (NaCl). A broad range of data was compiled from 142 experimental data points of different drilling mud samples in order to create and test the model. For training, the model employed 100 data points out of the total 142, while for testing, the remaining 42 were put into use. Results showed that the ANN-based model had exceptional prediction accuracy where model fitting to the training dataset produced a high correlation coefficient (R) of 0.99 and generalization to the testing dataset gave a more than satisfactory R value of 0.98. The results confirm that the ANN model proposed in this study is capable of predicting plastic viscosity with reasonable accuracy, providing a quick and efficient substitute to traditional laboratory testing procedures. This predictive model can be significantly helpful in dealing with real-time issues related to decisions made in the drilling operations, the design of mud, and problems with fluid viscosity in drilling.</p>