The chapter discusses the various traffic noise prediction models developed in India with an objective of traffic noise predictions and forecasting. These models had been developed based on various techniques such as Multiple Linear regression approach, Artificial Neural Networks, calibration of foreign models such as RLS90, FHWA, etc. and based on machine learning techniques. However, there is a need for a validated and reliable model suitable for carrying out the Environmental Impact Assessment (EIA) studies w.r.t noise that is also recognized by the state/central pollution control agency integrated with a GIS interface and suitable for devising the noise control measures. The study also discusses the various attributes of an efficacious model that can be indispensable for urban planning in order to optimize the road layouts and minimize noise impacts on residential and silence zone areas. The major point of debate among the scientific fraternity arises in developing either a source-specific model, or a regression-based model or using the advanced time-series techniques for traffic noise modeling.