With a modified microcytotoxicity assay, the effects of lymphocytes from normal volunteers and from patients with a variety of solid malignant tumors were tested in vitro against both normal and neoplastic target cells. The study was divided into three sequential time phases in each of which a different method of lymphocyte separation was used. Lymphocytes were separated by a Ludox-polyvinyl-pyrrolidone technique in the first phase, by nylon wool filtration in the second phase, and by a Ficoll-Hypaque technique in the third phase of the study. In the first two phases of the study, lymphocytes from both normal volunteers and cancer patients were frequently toxic on both tumor cells and fibroblasts. In the last phase of the study types of target cells. Cancer patient lymphocytes were more frequently toxic on tumor cells but no more frequently toxic on fibroblasts than were normal lymphocytes. The different results obtained in the last phase of the study cannot be attributed solely to the different method of lymphocyte separation, since other factors, such as better growth status of the target cells and greater facility in performing the assays, might also be responsible. Results of the last phase of the study raise the possibility that both specific and nospecific reactions may contribute to the toxicity observed in this laboratory with cancer patient lymphocytes tested against tumor cells. It has not yet been possible to reduce the level or frequency of the nonspecific reactions to such a degree that the microcytotoxicity assay can be used clinically to document clearly or to follow the putative tumor-specific immunity of individual cancer patients.