Proliferation of GIST cells is mostly driven by either KIT or PDGFRA mutation. Imatinib have shown high activities on advanced GISTs with KIT or PDGFRA mutations and its efficacy depends on the genotype. Primary resistance to imatinib was rare but principally seen in GIST without mutation in the both genes, or in those with resistant mutations such as KIT D816H/V and PDGFRA D842V. Even effective, there appears secondary resistance to imatinib with a median follow-up of 2 years. Secondary resistance is clinically diagnosed by enhanced CT with appearance of a nodule in a mass, tumor enlargement, and/or a new lesion. From pathological and molecular examinations, these resistant lesions show monoclonal development of resistant tumor cells. Causes of secondary resistance are mainly due to secondary mutations in the kinase domains of KIT or PDGFRA (70–80%), and rarely due to over-expression of KIT and/or increase in copy number of mutated KIT or PDGFRA (10%). Secondary mutations may be induced before and after imatinib from mathematical models and in vitro experiments. The others are postulated to be due to acquisition of new but unknown proliferation mechanisms (10%). Secondary mutations in the kinase domains are accompanied by concomitant re-activation of the corresponding tyrosine kinases even in the presence of imatinib. Secondary mutations were found on the same allele to the primary mutations, and GIST with imatinib-naïve KIT exon 11 mutations acquired more frequently secondary mutations than those with exon 9, relatively resistant mutation. Secondary mutations also have hot spots either in the ATP-biding domain including KIT exon 13 (V654A) and exon 14 (T670I), or in the activation loop of exon 17 such as, codons 809, 816, 820, 823 and 829, for instance. Imatinib-resistant GISTs with the former mutations in ATP-biding domain are sensitive to sunitinib, a second-line agent, whereas those with the latter mutations in the activation loop are resistant to sunitinib.