The biofragmentable anastomosis ring (BAR) has been used successfully for anastomoses from the stomach to the upper rectum. The healing of intrathoracic esophageal anastomoses performed with the BAR or an end-to-end anastomosis (EEA) stapler on an experimental model was compared. Parameters of tissue repair were evaluated: macroscopic examination, bursting strength (BS), collagen (hydroxyproline, or HP), histology (H&E and Picrosirius red staining for collagen). A series of 48 mongrel dogs were randomly separated into two groups (30 BAR, 18 stapler) and subgroups according to the time of autopsy (days 4, 7, 14, 28). Mortality was 13.3% (4 BAR cases) with two deaths not related to surgery (excluded). There were four leaks in the BAR group (14.3%) and no leaks or deaths but two strictures in the stapler group. BS was significantly higher in the BAR group during the first week, and values were almost equal from the second week with both methods. The HP rate was significantly reduced on days 4 and 7 in both groups compared to the reference values; the values were close to reference values from the second week (lower in the BAR group). Stapled anastomoses caused less pronounced inflammation and were associated with an earlier start of regeneration, but the difference was not significant compared to that in the BAR group. Accumulation of new collagen (green polarization) started on day 7 in both groups, but maturation (orange-red polarization) was significantly more advanced in the BAR group after the second week. A strong linear correlation between the BS and HP rate was found with both methods. There was no significant difference in the complication rate or healing of intrathoracic BAR and stapled anastomoses. The BAR method is simple, quick, and safe; and it seems to be a feasible procedure for creating intrathoracic esophageal anastomoses in dogs.