Abstract Background Anastomotic leakage rates remain unacceptably high, warranting reconsideration of current anastomotic technique. Anastomotic healing may improve by abrading the serosal surface of bowel ends that are invertedly anastomosed, based on the concept that serosal damage evokes inflammatory adherent processes. It is studied if local abrasion leads to stronger anastomoses and reduces leakage. Methods Ninety-eight Wistar rats were allocated to six groups. Either a regular anastomosis (RA) or abraded anastomosis (AA) was constructed in the proximal colon. Animals were sacrificed at day 3 (groups RA3 and AA3, n = 2 × 17) or day 5 (groups RA5 and AA5, n = 2 × 17). Groups RA-Dic and AA-Dic (n = 2 × 15) received diclofenac from day 0 until sacrifice on day 3 to impair anastomotic healing. Outcomes were leakage, bursting pressure, breaking strength, adhesions, and histological appearance. Results Both in abraded (AA3 and AA5) and control (RA3 and RA5) groups without diclofenac, 1 of 17 anastomoses leaked (6%). Leak rate was 9 of 15 (60%) in group AA-Dic and 8 of 15 (53%) in RA-Dic (P = 1.0). The bursting pressure in group RA3 (127 ± 44 mm Hg) was higher (P = 0.006) compared with group AA3 (82 ± 34 mm Hg), breaking strength was comparable (P = 0.331). Mechanical strength was similar between groups RA5 and AA5. Abrasion did not increase mechanical strength in the diclofenac groups. Adhesion formation was not different between groups. Histology showed dense interserosal scar formation in abraded groups, compared with loose connective tissue in control anastomoses. Conclusions Abrasion of serosal edges of large bowel ends invertedly anastomosed does not improve anastomotic strength, neither does it reduce leakage in anastomoses compromised by diclofenac.