Abstract: In order to realize adaptive compensation for positioning error of precision measurement platform, a new method has been proposed based on monitoring measurement environment and support vector regression for ensuring the high-precision of measurement platform used for optical element in different measurement environment. Firstly, a prediction model was established with the theory of support vector regression (SVR), using several group data of temperature, humidity, atmospheric pressure in different environment for predicting the maximum positioning error value. Then SVR was used iteratively to predict the positioning error value of any position, using both maximum positioning error value and environmental factors. Finally, the positioning error was input to the controller for compensation. Instruments including Renishaw laser interferometer along with temperature and humidity sensors were applied to experiments with the precision measurement platform. Experimental results indicate that, comparing the prediction data with the measured data, the average of the absolute difference is 0.88 m, and the Pearsons square of correlation coefficient is 0.99. After adaptive compensation, the average of positioning error decreases to 1.4 m from 43 m. It concludes that this method is feasible and accurate.