Uniaxial compressive strength (UCS) of rock is a crucial parameter extensively applied in various rock engineering applications. The UCS test needs the preparation of specimens of standard size, which sometimes are challenging due to fractured rock or soft rock. Due to these difficulties, UCS of rock is often predicted indirectly using physical and simple index parameters. In the present study, an Artificial Neural Networks (ANNs) approach was used to predict the uniaxial compressive strength (UCS) of metamorphic rocks based on datasets of ultrasonic pulse velocities (P-wave velocity and S-wave velocity). The ANN model was developed in Python with ReLU, sigmoid activation function, and trained using the back-propagation learning technique. The datasets that were used for training and testing the ANNs comprise different metamorphic rocks such as gneiss, schist, quartzite, metabasic, metavolcanic, and amphibolite from the Himalayan region, Chotanagpur plateau, and Decan traps. The coefficient of determination (R2) and root mean square error (RMSE) were calculated to evaluate the predictive performance of the ANN model. The developed ANN model based on collected datasets showed R2 and RMSE of 0.85 and 10.9 MPa, respectively. The prediction performance of the ANNs model was also compared with the performance of the multi-regression method on the same datasets, and it was observed that the ANNs model is a better predictor than the multi-regression method (R2 = 0.63 and RMSE = 17.9 MPa). Thus, the ANN model can be applied for more precise prediction of the UCS for metamorphic rocks in various engineering applications.