It has been demonstrated that tumor-bearing animals elaborate a low molecular weight (< 10,OOO daltons) factor capable of inhibiting in vitro lymphocyte function. It was postulated that removal of this factor would have a favorable effect on host immune response that would translate into improved tumor control. A study was conducted in rabbits bearing the VX-2 carcinoma. Ultrafiltration (UF) was performed 10 days following 1V tumor inoculation. UF was achieved by passing blood through an Amicon Diafilter @ (molecular weight cutoff 10 kD) positioned between the arterial and venous cannulae after heparinization. Two plasma volumes of ultrafiltrate were removed with continuous saline replacement. Two groups of animals received the nonspecific immunoadjuvant, Detox, at time of therapy. Survival in the UF group (N = 9) was compared to untreated tumor-bearing animals (N = lo), sham-operated animals (N = 6), animals receiving Detox (N = 7), and animals receiving UF plus Detox (N = 6). UF imparted a survival advantage when compared to controls (mean 35 days vs. 25 days, P < .01). The sham group had survival identical to controls. Detox alone conferred minimal survival advantage (mean 29 days, P > .05). However, UF + Detox demonstrated maximal survival benefit (mean 40 days, P < .01). We conclude that UF is an effective anticancer modality in this preclinical model. This study suggests that efforts aimed at eliminating suppressor molecules in cancer patients may be of benefit, especially when combined with biological response modifiers such as Detox.