Objective To develop a traditional Chinese medicine (TCM) syndrome score scale for acute gastrointestinal injury (AGI) in sepsis, and to carry out its reliability and validity analyses and its clinical preliminary application. Methods ① According to the characteristics of intensive care unit (ICU) patients, combined with the understanding of etiology, pathogenesis and physical signs of TCM and literature search, a preliminary framework of scoring system for TCM syndromes of AGI in sepsis was constructed to carry out the scoring by this scale. ② After the scale and data were obtained, the analyses of split-half reliability (indicated by Guttman's split-half reliability of the a and b groups), test-retest reliability and the internal consistency reliability (expressed by the Cronbach's coefficient α) were carried out, and the structural validity and criterion validity were also analyzed. ③ The AGI patients were divided into two groups according to the 28-day survival and death conditions, and the AGI TCM syndrome score, acute physiology and chronic health evaluation Ⅱ (APACHE Ⅱ) score, sequential organ failure assessment (SOFA) score, and multiple organ dysfunction syndrome (MODS) score were compared between the two groups to determine the best cut-off point for survival analysis. Results ① The first draft of the septic AGI TCM syndrome rating scale was prepared, The TCM syndrome indicators include: abdominal distension, constipation/diarrhea, diet situation, vomiting/stomach retention, tongue proper, tongue coating, pulse manifestation, belching, body temperature, and accompanied syndrome, there were 6 points for scoring, 0 - 6 points, and they were divided into normal (0 points), mild (2 points), moderate (4 points), and severe (6 points) in severity. ② Eighty-eight patients with septic AGI were included in the final statistics. The retest of correlation coefficient of this scale was R = 0.974 (> 0.85), Guttman's split-half reliability was 0.793 (> 0.7) and the Cronbach's coefficient α was > 0.7. This scale was suitable for factor analysis. After rotation, 3 factors were determined, which were named as TCM syndrome differentiation, related physical signs, and gastrointestinal tolerance. After modeling, the confirmatory factor analysis showed that the model approximate error root mean square (RMSEA) was 0.07 (< 0.08), and the goodness of fit index (CFI) = 0.90; the Pearson correlation analyses between the criteria validity of APACHE Ⅱ, SOFA, MODS scores and TCM 1 score and TCM 2 score of this scale showed that the r values were 0.802 and 0.752, 0.524 and 0.519, 0.619 and 0.590, respectively, all P < 0.01. ③ Compared with the survival group, TCM score (33.73±5.95 vs. 37.28±5.26, t = 2.945, P = 0.004), the APACHE Ⅱ score (19.90±4.47 vs. 22.28±5.79, t = 2.069, P = 0.043), SOFA score (8.73±1.11 vs. 9.64±1.38, t = 3.329, P = 0.020) in the death group were significantly decreased; MODS score in the death group showed a decreasing trend (6.65±1.22 vs. 7.28±1.60, t = 2.078, P = 0.050). Cox regression analysis showed that when the survival analysis was performed with a cut-off point of 35, the 28-day survival rate of patients with TCM syndrome score ≥ 35 was significantly lower than that of patients with < 35 score, χ2= 6.362, P = 0.012. Conclusions The TCM syndrome rating scale for AGI in sepsis was successfully prepared. The statistical reliability and validity of this scale are good. Preliminary clinical application shows that this scale can predict the prognosis and severity of patients with septic AGI. Trial registration China Clinical Trial Registry Center, ChiCTR-IOR-15007625.