We have previously shown (1, 2) that mice with mammary tumors can always be identified by their very high plasma levels of gp52, a 52,000 mol wt glycoprotein of the mouse mammary tumor virus (MMTV). The present investigation demostrates that the tumor is the principal source of the plasma gp52 since surgical excision is invariably followed in the first 9 days by a sharp decreasing (10-100-fold) of the gp52 levels. Control animals in which the tumors were left in place by a "sham" surgical procedure maintained their high level of gp52, which continued to increase as the disease progressed. The behavior of the gp52 after surgical removal suggests that gp52 plasma concentrations are diagnostically and prognostically informative, as indicated by the following finding: (a) All tumor recurrences were correctly diagnosed by increases in gp52 levels, and some were detected 4-7 days before they were found by palpation. (b) Tumor regrowths were accompanied by continued increases in plasma gp52 concentrations at rates that usually matched the speed of tumor development. (c) The only animals that remained tumor free at the termination of the experiment were those that maintained their gp52 levels at or below 15 ng/ml. (d) The probability of a tumor-free animal relapsing within 2 wk is much higher if its gp52 level is above the mean. (e) More remarkably, the plasma levels of gp52 at the time of surgery are superior to the size of the tumors removed as prognostic indicators of eventual surgical "cures". The availability of a specific and sensitive systemic measure of disease status should augment the usefulness of the murine mammary tumor model by catalyzing a more rapid acquisition of information on the therapeutic effectiveness of the new and varied drug combinations being tested for adjuvant chemotherapy.