ABSTRACT: Aim: Only few efforts have been taken to investigate the potential existence of disease-specific differences in the safety profile of everolimus. We analyze here the correlation between different patient and tumor characteristics on the safety profile of this molecule. Information on treatment response is also provided. Methods: Consecutive patients with metastatic renal cell carcinoma (mRCC), pancreatic neuroendocrine tumors (pNET) or biliary tract cancer were included in this retrospective study. All patients received everolimus 10 mg/day or 5 mg/day. Clinical assessments were performed every 3 weeks. Results: In total, 98 patients were enrolled: 51 with mRCC, 25 with pNET and 22 with biliary tract cancer. The incidence of toxicities (any grade) was 76% with mRCC, 64% with pNET and 95% with biliary tract cancer. Patients with biliary tract cancer also presented a higher frequency of severe toxicities: 64 versus 18% with mRCC and 32% with pNET. Multivariate analysis disclosed that biliary tract cancer (odds ratio [OR]: 23.8; 95% CI: 6.0–117.8; p < 0.0001) is a predictive factor for the development of toxicities during everolimus treatment. No correlations between liver metastasis and toxicities were identified. Disease control rate (DCR) was 45% in mRCC patients, 96% in pNET and 50% for biliary tract cancer patients. pNET tumors were associated with a higher DCR than the mRCC and biliary tract cancer (OR vs mRCC: 66.7; 95% CI: 6.2–276.5; p = 0.004; OR vs biliary tract cancer: 2.6; 95% CI: 0.5–14.2; p = 0.025). Conclusion: This study suggests that the safety profile of everolimus is acceptable in patients with either mRCC or pNET. In addition, the onset of toxicities is associated with an improved DCR. In patients with biliary tract cancer, everolimus is safe but associated with a higher incidence of adverse events.