An objective review of the experimental and clinical literature comparing stapled- and hand-sutured anastomoses in gastrointestinal surgery is presented. Experimental studies support the hypothesis that stapled-sutured lines heal by primary intention, thus avoiding the inflammatory time lag, and resulting in a stronger wound during the first post-operative week. In contrast, manual sutures always produce an early inflammatory reaction with a transitory rise in collagenase activity in the submucosa, which explains the particular fragility of the anastomosis during that phase. Nevertheless, it was difficult to find a concordance between this histological quality of the stapled suture line and clinical results. In oesophageal surgery, a slightly higher rate of stenosis and a lower rate of anastomotic leakage were observed. In gastric and colorectal surgery, the superiority of the stapled suture was not proved. However, in colorectal randomized studies, the presence of "withdrawn" group for which the stapler was the only technically possible alternative, provides undeniable proof that this technique is an indispensable adjunct in modern digestive surgery. Conversely, its value seems negligible in surgery of the small bowel or of the intraperitoneal colon where manual suturing remains the technique of choice.