In order to explore the respective role of wound healing and suture material on metachronous carcinogenesis at colonic lines in rats, tumour yield was studied after the administration of a chemical carcinogen (azoxymethane) at the moment that in one group of rats most of the suture material was still present and crypt cell proliferation elevated, while in another group, no more suture material was present and crypt cell proliferation rate (CCPR) normalized at the anastomotic site. Azoxymethane (15 mg/kg/week, s.c., during 6 weeks) was administered in male Sprague-Dawley rats (n = 105) 8 weeks after the creation of an anastomosis in the ascending and descending colon with either stainless steel sutures (group A, n = 30) or fast-absorption Vicryl (Vicryl Rapide, group B, n = 30). A control group (group C, n = 30) underwent a sham laparotomy before the administration of azoxymethane, while the animals of a fourth group (group D, n = 15) were not operated upon and received no azoxymethane. Twenty-six weeks after the first injection of azoxymethane there was no significant difference in the total colorectal tumour yield in the three operated groups (A, B, C), but a significantly greater proportion of anastomotic tumours (28/68 vs. 13/88, p < 0.01) and more anastomotic tumours per rat (28/23 vs. 13/28, p = 0.01) as well as more rats with anastomotic tumours (16/23 vs. 11/28, p = 0.04) were found in the steel-sutured group (A), compared with the control group (C).(ABSTRACT TRUNCATED AT 250 WORDS)