Soluble sugars and starches are important metabolites of plant life and physiological markers of plant stress response. There is an urgent need to develop a non-destructive and rapid method for determining plant starch and soluble sugar contents. Electrical impedance tomography (EIT) technology has been used to determine the physiological state and cold resistance of select plant tissues. However, so far there have been no reports on the use of EIT for the rapid estimation of soluble sugar and starch contents. In this study, EIT was used to obtain reconstructed voltage values and estimate starch and soluble sugar contents in the stems of three Rosa hybrida cultivars during February to May, which were grown in the Specimen Park (38° 50′ N, 115° 26′ E) of Hebei Agricultural University, Baoding City, Hebei Province, China. Stems from two of the cultivars were used for establishing regression models for starch and soluble sugar contents as functions of reconstructed voltage values. The third cultivar was used to test the accuracy of the regression models. The quadratic regression model was best for determining soluble sugar content and the logarithmic regression model was best for determining starch content. Thus, this research provided technical support for using EIT to analyze changes in physiological parameters and to rapidly estimate physiological indexes of plants. More studies were now needed to validate the results in this paper.