The purpose of this study was to determine whether the characteristics of compression anastomoses created by a new device are comparable to existing stapler technology. A total of 18 pigs were studied, and each served as its own control using a 27-mm compression device and a 29-mm stapler. Anastomoses were randomized to proximal and distal positions along the rectum and were separated by 10 cm. Six nonsurvival pigs were sacrificed at zero time to failure test the anastomoses. Twelve pigs were sacrificed at 14 days. Failure pressures, circumferences, and radiographic leak rates were determined. Anastomotic tissue was processed for matrix metalloproteinase, collagen, and elastin levels. The compression anastomoses had higher mean failure pressures than stapled anastomoses at zero time (103 vs 29.9 mm Hg). At 2 weeks, there was no difference between failure pressures (256 vs 250 mm Hg). During burst testing, 3 of the compression anastomoses failed at the anastomosis at 2 weeks, whereas none of the stapled anastomoses failed. The mean anastomotic circumference of the compression anastomoses was narrower than the stapled anastomoses (9.63 vs 11.25 cm, P = .001). There were no clinical leaks or radiographic leaks by barium enema at 2 weeks. There was no difference between matrix metalloproteinase, collagen, or elastin content based on tissue analysis. There were dense adhesions to 7 of 12 (58.3%) of the stapled anastomoses, whereas only 1 of 12 (8.3%) of the compression anastomoses had flimsy adhesions. A new compression anastomosis technique using a nickel-titanium alloy may be an advance in technology by reducing leaks and eliminating foreign material in the anastomosis.