The development of roads has been one of the nation’s most essential infrastructural initiatives. It is an essential mode of transportation that plays an important role in our everyday lives. Because of its importance, the government has allotted large budgets in making roads in different parts of the country. The quantity and complexity of road construction projects have substantially expanded in recent years. Numerous novel methods and technology have been developed to facilitate road construction budgeting, planning, and decision-making. Using Artificial Neural Network (ANN), this study constructed a forecasting model to accurately anticipate the future costs of road improvements. Between 2017 and 2020, fifty (50) completed road projects from the Department of Public Works and Highways (DPWH) Regional Office XI were utilized by the researcher. The DPWH RO XI is one of the country’s largest implementing offices for constructing public roads catering the entire Davao Region. This research used the project cost as the dependent variable while the independent variables are the construction activities’ revised duration and variation order in running the model. Multiple linear regression model performance was compared to the performance of the neural network prediction model. The data included the major construction activities for road projects with its corresponding revised duration, actual and planned cost, and the reason for variation order. It demonstrates that the neural network models outperform to the multiple linear regression (MLR) model in terms of prediction accuracy. This research offers a model to the government agencies and contractors implementing road construction in predicting road construction costs more accurately.