Objectives Evaluation of surgical and non-surgical air-polishing in vitro efficacy for implant surface decontamination. Material and methods One hundred eighty implants were distributed to three differently angulated bone defect models (30°, 60°, 90°). Biofilm was imitated using indelible red color. Sixty implants were used for each defect, 20 of which were air-polished with three different types of glycine air powder abrasion (GAPA1-3) combinations. Within 20 equally air-polished implants, a surgical and non-surgical (with/without mucosa mask) procedure were simulated. All implants were photographed to determine the uncleaned surface. Changes in surface morphology were assessed using scanning electron micrographs (SEM). Results Cleaning efficacy did not show any significant differences between GAPA1-3 for surgical and non-surgical application. Within a cleaning method significant (p < 0.001) differences for GAPA2 between 30°(11.77 ± 2.73%) and 90°(7.25 ± 1.42%) in the non-surgical and 30°(8.26 ± 1.02%) and 60°(5.02 ± 0.84%) in the surgical simulation occurred. The surgical use of air-polishing (6.68 ± 1.66%) was significantly superior (p < 0.001) to the non-surgical (10.13 ± 2.75%). SEM micrographs showed no surface damages after use of GAPA. Conclusions Air-polishing is an efficient, surface protective method for surgical and non-surgical implant surface decontamination in this in vitro model. No method resulted in a complete cleaning of the implant surface. Clinical relevance Air-polishing appears to be promising for implant surface decontamination regardless of the device.