In clinical practice, there are often discrepancies between the oncological prognosis of gastrointestinal stromal tumors (GIST) and the actual clinical course. This study aimed to check with our collective how reliably the current classifications (Miettinen, Fletcher) predict the prognosis of GIST and to evaluate whether an extension of the classifications by the parameter proliferation activity could make sense. This prospective study enrolled 58 patients who underwent surgery on GIST from 01/2006 to 12/2016. The postoperative course (curation, recurrence, progress) was correlated with the identified risk classification and the proliferative activity. Coincidences with other tumors were strikingly common in patients with GIST (43%). Based on the risk group assignment of GIST, no assessment of the probability of the occurrence of second neoplasia could be derived. Individual patients were under- or over-graduated concerning the assessment of biological behavior based on the standard risk classifications. The inclusion of proliferative activity did not allow for a more precise predictive power – neither to the risk of recurrence and metastasis nor to the development of a second neoplasia. The study showed that there is currently no parameter or logarithm that reliably predicts the biological behavior of GIST. Due to the frequency of coincidence of second neoplasia and (rare) distant metastases, for everyday clinical practice, appropriate staging diagnostic and regular follow-up care should also be used for benign GIST.