S tudies of immunoglobulin synthesis, total body tumor cell number, and tumor kinetics were carried out in a series of patients with IgG multiple myeloma. The changes in tumor size associated with tumor growth or with regression were underestimated when the concentration of serum M-component was used as the sole index of tumor mass. Calculation of the total body M-component synthetic rate (corrected for concentration-dependent changes in IgG metabolism) and tumor cell number gave a more accurate and predictable estimate of changes in tumor size. Tumor growth and drug-induced tumor regression were found to follow Gompertzian kinetics, with progressive retardation of the rate of change of tumor size in both of these circumstances. This retardation effect, describable with a constant α, may be caused by a shift in the proportion of tumor cells in the proliferative cycle. Drug sensitivity of the tumor could be described quantitatively with a calculation of B_O, the tumor's initial sensitivity to a given drug regimen. Of particular clinical significance, the magnitude of a given patient's tumor regression could be predicted from the ratio of B_O to α. Mathematical proof was obtained that the retardation constant determined during tumor regression also applied to the earlier period of tumor growth, and this constant was used to reconstruct the preclinical history of disease. In the average patient, fewer than 5 yr elapse from the initial tumor cell doubling to its clinical presentation with from 10¹¹ to more than 10¹² myeloma cells in the body. The reduction in total body tumor mass in most patients responding to therapy ranges from less than one to almost two orders of magnitude. Application of predictive kinetic analysis to the design of sequential drug regimens may lead to further improvement in the treatment of multiple myeloma and other tumors with similar growth characteristics.