In many transplant centers, the T lymphocytes of transplant patients are routinely monitored, both to predict and diagnose the cellular events that result from transplantation and to evaluate the effectiveness of immunosuppressive therapy. During the course of this monitoring, we have observed anomalous results: the number of T cells said to be present varies greatly depending on the surface marker used for the assay. It is shown, in this study, that these "anomalous" results can be predicted from a knowledge of the spectrum of specificity of the anti-lymphocyte antibody therapy being administered. Because polyclonal antibodies have varied and unstated specificities, it is difficult to interpret results obtained from monitoring T-cell numbers or subset ratios without some knowledge of the specificities of the drugs being used.