The increase in fast moving traffic and narrow roads are causing vehicle congestion which results in vehicles moving in low gears, accelerating, stopping and horn blowing, etc. All these are the contributing factors for considerable noise in industrial towns in India. The percentage of heavy vehicles in the traffic stream is also a major factor contributing to high noise levels. This paper delineates the systematic noise investigation in Dhanbad, an industrial town in the coal belt. Out of the six monitoring stations within Dhanbad town – Bank More, Shramik Chawk and Court More stations show maximum LAeq levels [79.0 to 84.1 dB] due to higher number of vehicles. On the other hand, at G.T. Road at Govindpur free flowing traffic outside Dhanbad town registers maximum LAeq[81.0 dB] during evening hour (7.00p.m to 8.00p.m) due to the greater number of heavy vehicles (189). The noise generated within Dhanbad town is more than that of the free flowing highway at Govindpur G.T. Road. A car in general generates 15 -dB (A) less noise than a heavy commercial vehicle. It has also been observed that sudden braking followed by gearing acceleration leads to an increase of 10– 15 dB (A) noise level. The frequency spectrum analysis reveals that the highway (G.T. Road) noise has a well-defined sound pressure level (SPL) at 63 Hz., representing the firing frequency of the vehicles at high speed. Again, a comparatively less alarming frequency at about 1 kHz may be due to the combined effect of tyres and wind. The alarming SPL for the traffic inside town has been in the range of 63 Hz to 125 Hz due to the lower speed of the vehicles. After that, there is no peak at higher frequency range, indicating that at relatively low speed tyres and wind make a negligible contribution to the noise spectrum. The Noise Ratings (NR), which represent the annoyance factor, are observed varying from morning to evening. In most of places, NR during the morning session show peak value due to prevalence of dominant high frequency. Statistical models / relationships between LAeq, total traffic volume per hour (Q) and percentage of heavy vehicles over total vehicles (P) are developed which can be used for the prediction of traffic noise in similar situations. The authors stress the need of curbing the traffic noise by taking appropriate steps to bring about a smoother traffic flow at moderate speeds and minimum horn blowing. The necessary steps will include widening of roads, diverting the traffic by means of some bypass roads, making traffic one-way, removing the business congestion in the roadway, avoiding or removing inter sections at near by places, etc. Besides, provision of suitable barriers in the sensitive zone of the roadways can go a long war way to reducing the noise situation.