One hundred and sixty-seven patients undergoing laparotomy were randomly allocated to a continuous layered closure technique with absorbable monofilament polyglyconate (PG) (Maxon*) or non-absorbable polyamide (PA) (Ethilon*). Laparotomy wounds were closely observed during the postoperative hospital course and all patients were reviewed at one month, six months and one year. Any wound complications were noted. Two patients in the PA group (3.0%) presented with burst abdomen and one (1.9%) in the PG group (ns). The incidence of incisional hernia was not statistically different between the two groups (4/64, 6.0% in PA group and 4/65, 6.2% in PG group). The postoperative wound infection rate was 21.0% in the Pa group and 15.4% in the PG group (ns). The present study clearly shows a major increase in incidence of wound failure in patients with infected abdominal wounds in both groups (28.6% vs 3.8% in PA group p less than 0.025; 20% vs 5.5% in Pg group p less than 0.05). This clinical trial confirms the important roles of wound infection and respiratory failure in the occurrence of wound failure; the choice of an absorbable or a non-absorbable suture material seems to play a minor role only.