We employed a tumor-type-specific tissue culture model utilizing three human colon carcinoma cell lines to (i) assess the effects of schedule, sequence, dose, and duration of exposure on the antiproliferative activity of combinations of tumor necrosis factor (TNF) and interferon-gamma (IFN-gamma) and to (ii) determine the effects of this combination on the production of a TNF-inducible protein, IFN-beta 2, and an IFN-inducible protein, p78. A statistical model was developed to ascertain the effects of each of three of the variables (sequence, dose, and duration) on the other two. With minor exceptions, the maximal antiproliferative effect in all three cell lines was observed when IFN-gamma and TNF were administered simultaneously, regardless of the doses of each agent and duration of exposure. Exposing the cells to TNF before IFN-gamma resulted in the least inhibition in cell growth. In all three cell lines, the antiproliferative effects of each treatment group were related directly to the duration of exposure. In two of the three cell lines, an intermittent schedule was as effective as a 24-h exposure. No p78 induction was observed in the HCT 116 cell line with IFN-gamma alone, TNF alone, or the combination; p78 was slightly induced in the SKC01 and VACO 9P cell lines with IFN-gamma alone, and was synergistically induced by the combination of TNF and IFN-gamma. Treatment with a neutralizing antibody to IFN-beta did not reverse the antiproliferative effect of any of the three treatment groups in HCT 116 cells. We conclude that the maximum antiproliferative effect in human colon carcinoma can be achieved by the prolonged simultaneous administration of high concentrations of each drug. If clinical toxicity prohibits this schedule, an intermittent schedule of administration may be effective. The mechanism of the synergistic effect is not due to the induction of IFN-beta or the p78 protein.