Aim: To develop an online tool based on an evidence-based predictive model, which allows clinicians to accurately predict the risk of peri-implantitis in candidates for dental implant therapy. Material & methods: A retrospective study of patients attending a university implant review clinic was performed. The presence of peri-implantitis and related risk factors were recorded. A predictive model for peri-implantitis was then developed based on this data. Results: 460 patients having 1,432 implants were included. Peri-implantitis was found in 78 (17%) patients. For partially edentulous patients (n=350, 60% female, average age 64.1 years), susceptibility to periodontitis (OR 0.48 [0.24;0.94], p = 0.03), the number of sites with probing pocket depth ³ 5 mm (OR 0.2 [0.10;0.40], p < 0.01) and smoking (OR 0.25 [0.09;0.66], p < 0.01) were significantly associated with peri-implantitis. For fully edentulous patients (n=50, 72% female, average age 72.2 years), implants placed in the maxilla displayed a greater risk (OR 0.15 [0.02;0.87], p = 0.03) of developing periimplantitis. A predictive model for the development of peri-implantitis was created, based on 8 patient-related risk factors for partially edentulous patients (sensitivity = 90.2%, specificity = 55.0%) and 4 risk factors for fully edentulous patients (sensitivity = 100%, specificity = 51.3%). Conclusions: The predictive model can be used for a pre-operative risk assessment of partially edentulous patients. Further validation and refinement of the model with additional data could enable its use for fully edentulous patients, and will improve its predictive power, thereby increasing its reliability.