Background: There is no consensus regarding the best management strategy for diagnosing and treating GI stromal tumors (GISTs). Objective: Our purpose was to examine the practice patterns of endosonographers in diagnosing and managing GISTs, particularly features of GISTs suggestive of malignancy, features that prompt surgical referral, and surveillance patterns. Design: An invitation to complete an online survey was e-mailed to all 413 members of the American Society for Gastrointestinal Endoscopy EUS Special Interest Group. Results: A total of 134 (32%) members responded; 59% of respondents use EUS features combined with FNA findings to diagnose GIST, and 89% consider a c-kit-positive stain on FNA most suggestive of GIST. However, 60% would diagnose GIST when cytologic samples are insufficient for diagnosis, and 40% would diagnose GIST if cytologic samples are sufficient but c-kit is negative. A total of 92% use size as the main criterion to distinguish benign from malignant GISTs, and 90% refer lesions >5 cm for surgery. For lesions not resected, 70% survey annually, 19% less than annually, 10% more than annually, and 1% do not survey. Limitations: The opinions of the respondents do not necessarily reflect the opinions and practices of endosonographers nationwide. There are inherent limitations to an online multiple-choice survey, including low response rates. Conclusions: There are substantial practice variations in diagnosing, resecting, and surveying GISTs. A majority of our survey respondents have made the diagnosis of GIST without FNA confirmation. Size >5 cm is the feature used most to predict malignancy and to prompt surgical referral. Surveillance practices for unresected GISTs are variable. Evidence is needed to establish practice guidelines in this area.