In this experimental work, the prediction models have been developed to predict acceleration amplitude tool vibration in hard turning. The material used for this experiment is austempered ductile iron grade 3. On-line acceleration amplitude of tool vibration in main cutting force and radial cutting force are measured. The cutting parameters are the depth of cut, feed rate and cutting speed in three levels. The experiments are conducted as per Taguchi L9 orthogonal array. ANOVA is used to calculate significant factors affecting the tool vibration amplitude. Linear square regression is used to estimate the acceleration amplitude of tool vibration main cutting force and radial cutting force. Based on the ANOVA result, it found that cutting speed and depth of cut affecting the acceleration of tool vibration in main cutting force and radial cutting force respectively. The developed regression model was verified by confirmation experiment. The estimated and experimental values for acceleration amplitude of tool vibration in main and radial cutting force adhere very much to 14.5 % and 15 % of maximum deviations respectively.