Gastrointestinal stromal tumours (GISTs) are the most common mesenchymal tumours of the gastrointestinal tract. Recent research has shown that small intestinal GISTs exhibit more aggressive features than gastric GISTs. To compare the clinical features of gastric and small intestinal GISTs for the further prediction of different prognoses, we conducted a retrospective study. 43 patients in the small intestine group and 97 in the gastric group were collected between January 2016 and December 2017. Data on demographics, preoperative lab results, clinicopathological results and surgical management were compared between groups. Significant elements were subsequently included in logistic regression analysis for further identification. The Kaplan-Meier method and log-rank test were used to calculate the relapse-free survival (RFS) rate and cumulative survival rate. Univariable analysis demonstrated that underlying disease, gastrointestinal (GI) bleeding, lymphocyte count, haemoglobin (Hb), albumin (ALB), platelet-to-lymphocyte ratio (PLR), thrombin time (TT), National Institutes of Health (NIH) category, Dog1, surgical procedure types and postoperative hospitalization were different between the two groups. Among these factors, logistic regression analysis identified that patients in small intestinal group exhibited significantly higher GI bleeding rate (p = 0.022), NIH category (p = 0.031), longer postoperative hospitalization time (p = 0.001) with lower TT value (p = 0.030) than those in gastric group. The log-rank test indicated that the location of the GIST (p = 0.022), GIST with GI bleeding (p = 0.027) and NIH category (p = 0.031) were independent prognostic predictors for poor outcome regarding RFS. Regarding cumulative survival, only the location of the GIST (p = 0.027) was an independent prognostic predictor for poor outcome. Thus, we concluded that small intestine GISTs were associated with lower TT, recurrent GI bleeding, advanced NIH category and extended postoperative hospitalization. Nevertheless, future multicentre prospective study are expected to validate our results.