The effect of surgical removal of "primary" tumors on the cytokinetics of local tumor remnants, secondary implants, and metastases was investigated in three different rat tumor models in the Wag/Rij rat: a slow-growing (MCR83) and a fast-growing (EMR86) hormone-dependent mammary tumor and a rapidly, but autonomously growing carcinoma (MCR86). The latter two tumors had metastatic potential. Cell kinetic studies were done using in vivo labeling with 5'-bromodeoxyuridine (BrdUrd). Thirty-three hours after removal of a subcutaneous MCR83 flank tumor, secondary implants showed a significant (P less than 0.05) but transient increase in the BrdUrd labeling index (LI). A more rapid and prolonged increase, lasting for at least 7 days, was observed in EMR86 lymph node and lung metastases. In both models, no effect was observed after sham surgery (consisting of opening and closing of the skin under anesthesia). Removal of MCR86 tumors (growing in the hind leg muscle) also resulted in a rapid, transient LI increase in metastases. Continuous BrdUrd labeling experiments in this tumor model did not favor the hypothesis that the LI increase predominantly resulted from an increase in the growth fraction. Moreover, in this model, the effect was related to operation trauma. A similar increase in LI, although smaller than after tumor removal, was seen after major surgical trauma in MCR83 flank tumors. These results indicate that in the rat, tumor removal and/or major surgical trauma may modulate the cytokinetics of distant metastases significantly. A study of the systemic, possibly endocrine, factors involved in the growth-stimulating effect of surgical trauma in these rat tumor models may help to assess the clinical relevance of these findings for patients with breast cancer.