The problem of audio signals based road traffic density estimation is consider in this paper. The vehicles produces various noises(sounds), depending on it's speed, load, and it's mechanical condition, the noise produces such as tire noise, engine noise, engine-idling noise, air-turbulence noise and occasional honks. In the various signals there are various types of spectral content which are different from each other hence classify between them in three classes such as jammed flow traffic(0-10km/h), medium flow(10-40km/h) and free flow(40km/h and above). Based on these extract the features from acoustic signals using Grey level co-occurrence matrix(GLCM) use a discriminative classifier such as a support vector machine(SVM) classifier Which results in high accuracy.