Imatinib resistance is the most important clinical issue in patients with gastrointestinal stromal tumor (GIST). However, the association of imatinib resistance with the genetic characteristics of GIST has not been clearly defined. Our meta-analysis aimed to investigate the association between imatinib resistance and KIT and PDGFRA mutations in GIST. METHODS. We identified all relevant studies in PubMed and Embase. The effect sizes were calculated as prevalence or odds ratio (OR) with a random-effects model. RESULTS. We identified 10 eligible studies that included 1083 GIST cases. Total imatinib resistance was found in 35.5 % of PDGFRA-mutant tumors (OR = 2.9, P = 0.038), 33.7% of wild-type tumors (KIT and PDGFRA non-mutant tumors; OR = 2.8, P = 0.002), and 27.4% of KIT-mutant tumors (OR = 0.3, P = 0.001). Primary imatinib resistance was found in 50.0% of PDGFRA-mutant tumors (OR = 10.9, P = 0.031), 33.4% of wild-type tumors (OR = 5.9, P = 0.060), and 8.9% of KIT-mutant tumors (OR = 0.2, P = 0.025). KIT exon 9-mutant tumors showed primary resistance more frequently than exon 11-mutant and other tumors (OR = 7.6, P < 0.001). Regarding secondary resistance associated with KIT second-site mutations, the exon 17 mutation (54.5%) was most frequent, followed by exon 13 (38.3%) and 14 (13.4%) mutations. CONCLUSION. Our meta-analysis indicates that imatinib resistance is closely associated with KIT and PDGFRA genotypes in GIST. Thus, the mutational status of KIT and PDGFRA might predict response to imatinib in GIST patients.