In order to improve the TDR soil moisture sensor’s measurement accuracy, the impact of fertilization on the measurement accuracy was analyzed and appropriate calibration methods were proposed. Taking the TDR soil moisture sensor as the research object, three factors of soil texture, soil moisture content and fertilization amount were considered. Loamy sand in Kashgar, silty loam in Karamay and silty clay loam in Changji which are the main soil types in the cotton producing areas of Xinjiang were selected, five soil moisture levels and five fertilization levels were set for the three different types of soil, accounting for a total of 75 treatments. Then, a calibration model for TDR soil moisture sensor was established based on the drying method and the simulation results were validated. The results show that under different soil moisture conditions, the error of the TDR sensor increased with the increase of moisture content. At the same water content, the error of the TDR sensor under the condition of loamy sand increased with the increase of fertilizer application. The error of the TDR sensor under the condition of silty loam decreased first and then increased sharply with the increase of fertilizer amount. Under the condition of silty clay loam, the error of the TDR sensor increased first, then decreased and then increased with the increase of fertilization amount. A multivariate prediction model was established for the output value of TDR soil moisture sensor under experimental conditions, which was then calibrated by logistic model. The average absolute errors after calibration were -0.18%, 1.46%, and 0.01%, respectively. This indicates that considering changes in fertilization and moisture content is conducive to improving the measurement accuracy of TDR sensors.