Abstract: Abstract: In order to test the feasibility of diagnosing water deficiency of sugarcane using parameters related with stem diameter variations, both pot experiment and field experiment were conducted. In the pot experiment, pots were selected with 60 cm in length and width, and 40 cm in depth. Sugarcane t22 was planted in pots and field. Stem diameter variations were continuously measured. The diurnal variations of stem diameter at different growing stages under different water deficits were investigated, and maximum daily shrinkage (MDS) and daily increase (DI) were calculated based on relative variations (RV) of stem. In addition, the correlations between MDS, DI (from pot experiment) and soil moisture were analyzed. A water deficiency diagnosis model was established and assessed by the F test. The results showed that RV values were significantly (P＜0.01) different among different soil moisture treatments during the stem growing period include of seedling, tillering and elongating stage. The overall growth of sugarcane stem diameter was reduced by about 41.7% to 100% as soil moisture was less than 60% of field capacity. Elongating stage was most sensitive to soil moisture deficiency, followed by tillering stage and seedling stage. At the stem mature period (that was sugarcane's mature stage), the total amount of stem diameter growth (sugarcane planted in pots) was 0.049, 0.055, 0.048 and 0.033 mm in 16 observation days under conditions of normal irrigation, light drought, moderate drought and severe drought, respectively. The growth of stem diameter basically stopped at this stage and thus it was the least sensitive to water deficits. The regression analysis showed that MDS and DI were significantly correlated (P＜0.01) with soil water content at different growing stages of stem except for DI at mature stage, the coefficient of determination was 0.501?0.765. During the stem growing period include seedling, tillering and elongating stage, MDS and DI both had high response to water stress, and hence could be reliable indicators for diagnosing water status in sugarcane. MDS was more reliable at seedling and elongating stage, while DI was better indicators at tillering stage from the results of the coefficient of determination and root mean squared error. At the mature stages, MDS was more reliable due to its high sensitivity to water deficits. The upper and lower values of soil moisture were inputted as variable values into the regression equation, and thereby the threshold range of MDS and DI under water stress was determined. At the seedling stage, the MDS of 0.026?0.052 mm or DI of 0.087?0.160 mm indicated the soil was in a moderate drought condition, and the MDS of 0.052?0.078 mm or DI of 0.012? 0.087 mm indicated the soil was in a severe drought condition. At the tillering stage, the MDS of 0.046?0.099 mm or DI of 0.256?0.111 mm indicated the soil was in a moderate drought condition, and the MDS of 0.099? 0.153 mm or DI of -0.038?0. 111 mm indicated the soil was in a severe drought condition. At the elongating stage, the MDS of 0.082?0.174 mm or DI of -0.013?0.189 mm indicated the soil was in a moderate drought condition, and the MDS of 0.174?0.266 mm or DI of -0.216?-0.013 mm indicated the soil was in a severe drought condition. At the mature stage, the MDS of 0.072?0.111 mm indicated the soil was in a severe drought condition. The F test indicated that the established models could reliably predict water deficiency, and the coefficient of determination for prediction was between 0.515?0.787. The determined indicator, as well as the established water deficiency diagnosis models, is valuable for studies on irrigation scheduling for sugarcane based on stem diameter variations.