Summary Intracardiac injection, in hooded Lister rats, of syngeneic MC28 sarcoma cells never induced tumour growth in normal bowel. Tumour growth occurred at the site of a colonic anastomosis if surgery preceded tumour injection but not if it followed tumour injection, even by as little as 1h. Maximum enhancement of tumour growth occurred when the healing process had progressed between 2 and 8 days, with a peak at 5 to 7 days. The enhancing effect was largely over by the time the healing had progressed 14 days. The syngeneic OES5 breast carcinoma also grew at colonic anastomoses when surgery preceded tumour injection by 5 days, but not in normal colon. The MC28 sarcoma also grew at ileal anastomoses but not in the normal ileum after intracardiac injection. By injecting radiolabelled sarcoma cells, an estimate of the probability of a single bloodborne tumour cell lodging at a colonic anastomosis and leading to a tumour deposit was calculated to be of the order of 1:43 whereas the probability of the cell lodging in normal colon and causing a deposit is < 1:4 x 104.When cells from the transplantable syngeneic sarcoma and carcinoma used in this study are injected into the left ventricle of a rat, they distribute to all organs in proportion to the fraction of the cardiac output they receive (Murphy et al., 1986). However, some organs (e.g., adrenals and bone), commonly develop deposits, other organs (e.g., skin and lungs), occasionally develop deposits, whilst others (e.g., spleen and intestines), never develop deposits. This effect is reproducible with different tumours and appears to be a feature of the behaviour of the recipient tissue rather than the tumour (Murphy et al., 1986). The resistance of the colon to growth of these experimental tumours is paralleled by the clinical observation that the large bowel is a rare site for bloodborne secondary deposits from primary malignancies elsewhere in the body.Trauma to a tissue is known to enhance the ability of that tissue to support growth of tumour either from locally implanted cells (Jones & Rous, 1914), or from cells that reached the site of injury via the circulation (Robinson & Hoppe, 1962;Alexander & Altemeier, 1964;Fisher & Fisher, 1965).Following intracardiac injection of the two tumours used in this investigation, Murphy et al. (1988) found that growth occurred much more readily in healing laparotomy wounds than in the surrounding normal skeletal muscle. There have been no studies to investigate the effect of surgical trauma on the ability of the large bowel to support growth of tumour cells delivered by the circulation. The aims of this investigation were firstly to determine whether surgical trauma to the colon would enhance its ability to support growth of bloodborne cancer cells; and secondly, if enhancement did take place, to determine at which stage in the healing process the enhancement is at a maximum. experiments, and the OES5 breast carcinoma. MC28 is a methylcholanthrene-induced sarcoma and OES5 is an oestrogen-induced breast carcinoma (Senior et...