Abstract: For investigation of the effect of soil texture on the calibration of Time Domain Reflectometry (TDR), water content was determined gravimetrically for five textures. The experiments were carried out at the laboratory of Agriculture Collage of University of Kurdistan. For carry out of the experiments five textures and for each of them three repetitions were used. The studied textures were loamy sand, sandy loam, sandy clay-loam, clay loam and clay. The provided textures infuse inside the PVC cylinder with 30 cm height and 25 cm diameter. The vases were inside the water bath nearly 24 h. After 24 h of saturation, the soil water content of the samples was measured by oven-dry gravimetric and TDR methods. For each textures at any stage of measurement the determination of moisture and di-electric constant were carried out at three capture windows and with three repetitions. In each day the moisture measurement of each texture was carried out by these two methods until a moisture range was determined. The results of this research show that: (1) For all soil textures (loamy sand, sandy loam, sandy clay-loam, clay loam, clay) the measured moisture at the capture windows of 20 ns was more accuracy rather than other capture windows. (2) For all soil textures the TDR device can not measure the soil water content at capture window of 40 ns. (3) For loamy-sand and sandy-loam textures when the moisture was less than 20% and also for other textures for total range of measurement, the TDR device measured the moisture values less than the gravimetric method. This difference is increased when the soil ranged from light to heavy textures. (4) the comparison of the studied textures indicated that with increasing of clay content, the fitted lines slope is decreasing, but the RMSE values have increasing rate which it indicated the effect of clay content on the accuracy of TDR device. (5) From the statistical view at probably level of 1% there is not difference between the measured value of moisture by TDR device and gravimetric method. (6) For the all textures the fitted regression lines have high regression coefficient.