OBJECTIVES To assess the cleaning potential of commonly used implant debridement methods, simulating non-surgical peri-implantitis therapy in vitro. MATERIALS AND METHODS One-hundred-and-eighty dental implants were ink-stained and mounted in combined soft and hard tissue models, representing peri-implantitis defects with angulations of 30, 60, and 90° covered by a custom-made artificial mucosa. Implants were treated by a dental school graduate and a board-certified periodontist for 120 s with following instruments: Gracey curette, ultrasonic scaler, and an air powder abrasive device with a nozzle for sub-mucosal use utilizing glycine powder. All procedures were repeated 10 times for each instrumentation and defect morphology respectively. Images of the implant surface were taken. Areas with color remnants were planimetrically determined and their cumulative surface area was calculated. Results were tested for statistical differences using two-way anova and Bonferroni correction. Micro-morphologic surface changes were analyzed on scanning electron microscope (SEM) images. RESULTS The areas of uncleaned surfaces (%, mean ± standard deviations) for curettes, ultrasonic tips, and air abrasion accounted for 74.70 ± 4.89%, 66.95 ± 8.69% and 33.87 ± 12.59% respectively. The air powder abrasive device showed significantly better results for all defect angulations (P < 0.0001). SEM evaluation displayed considerable surface alterations after instrumentation with Gracey curettes and ultrasonic devices, whereas glycine powder did not result in any surface alterations. CONCLUSION A complete surface cleaning could not be achieved regardless of the instrumentation method applied. The air powder abrasive device showed a superior cleaning potential for all defect angulations with better results at wide defects.