Lipid nutrition effects were evaluated on the growth of a transplantable colon tumor (CT-26) at various sites in the BALB/c mouse. CT-26 implanted into the back or flank of these mice grew well independent of the quality or quantity of fat in the diet. However, when implanted in the mid-portion of the descending colon, tumor growth was related to the level of dietary saturated (coconut oil) or n-6 unsaturated (safflower oil) fat in the diet. Similar findings were obtained when the tumor was utilized in a pulmonary colonization assay. Dietary marine oil (mainly EPA, and DHA n-3 polyunsaturated oils) was found to markedly impair the growth of CT-26 implanted in the bowel and lung, but not in the back. Thus, CT-26 exhibits nutrition responsiveness at certain sites, but not at others. This may help to explain contradictory findings concerning dietary lipids in certain studies. Inhibition of tumor growth by marine oils may afford preventive or chemotherapeutic implications as its mode of action unfolds. Histologic findings in bowel tumors from mice fed marine oil but not other oils revealed focal areas of necrosis. It is appreciated that arachidonate metabolism is competitively interfered with by EPA in both cyclooxygenase and lipoxygenase pathways. The possibility is raised that the metabolism of marine oils in this model system may generate lipid peroxidation products to a greater extent than n-6 lipids and in turn is associated with focal areas of necrosis. A model system of nutritionally non-responsive and nutritionally responsive sites for the post-promotional growth of a bowel tumor affords the opportunity to explore lipid effects with control and test tumors in hosts fed identical lipid nutriture.