Purpose Objectives of the study are to analyze the correlation between [Ga-68]DOTATATE positron emission tomography (PET)/X-ray computed tomography (CT) measurements and various biological characteristics of gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs), and to determine optimal cutoff value of SUVmax (standard uptake value) to differentiate neuroendocrine tumors (NETs) and neuroendocrine cancers (NECs). Procedures Of the GEP-NEN cases (73 males, 53 females; age 18-77 years) with pathologically proven primary and/or metastatic lesions, 126 were studied. All of the short axes of lesions were larger than 0.5 cm in order to avoid the partial volume effect. Patients fasted for 6 h before the PET/CT scans. The dose of [Ga-68]DOTATATE was 100-200 MBq and the acquisition began at 1 h after injection. The lesion with the highest SUVmax in each patient was analyzed. Results In the total sample, the sensitivity of [68Ga]DOTATATE was 69.05 %. The sensitivities were significantly different among G1, G2, and G3 groups (72.22 %, 91.53 %, and 40.82 %, respectively; p < 0.01). The SUVmax of the G3 group was lowest. We also found that the sensitivity and SUVmax were significantly higher (p < 0.05) in patients with pancreatic NENs (Pan-NENs) than in patients with gastrointestinal NENs (Gi-NENs) and unknown primary NENs (Up-NENs). A significant negative correlation between SUVmax and Ki-67 was found (r = - 0.429, p < 0.01). Using SUVmax to differentiate neuroendocrine tumors (NETs) and neuroendocrine cancers (NECs), the area under the ROC curve (AUC) was 0.771 and the cutoff value of SUVmax was 11.25 (sensitivity 79.2 %, specificity 65.3 %). However, Pan-NENs did not show any statistical significance results in correlation and ROC analysis. Conclusion [Ga-68]DOTATATE PET/CT results showed a negative correlation with GEP-NEN cell proliferation and were complementary to Ki-67. Pan-NENs were different from Gi-NENs and Up-NENs when compared to somatostatin receptor expression.