Radionuclides provide biologically-distributed vehicles for radiotherapy of multifocal cancer. Two algorithms, fixed vs individualized, have been used to prescribe the therapeutic dose of radionuclide (GBq) for the patient. The individualized method for prescribing radionuclide dose takes variations in drug pharmacokinetics into consideration, whereas the fixed method depends, in part, on documentation that there is little interpatient pharmacokinetic variability for the radiolabeled drug. Two data bases, selected to compare iodine-131((131)I) and indium-111((111)In) labeled MAbs, were used to assess interpatient pharmacokinetic variability and its impact on radionuclide dose prescription. Pharmacokinetic data obtained over 7 days for non-Hodgkins lymphoma (NHL) patients given (131)I-Lym-1 (n = 46) or (111)In-Lym-1 (n = 13) were used to obtain cumulated activities. Although (131)I-Lym-1 often showed greater interpatient variability, (111)In-Lym-1 showed several-fold variability for many tissues. Both (131)I- and (111)In-Lym-1 had sufficient interpatient variability to be significant for radionuclide dose prescription, depending on the dose-limiting critical tissue. Interpatient variability exceeded intra- and interoperator variability and intrapatient variability over time for a single institution. In summary, the magnitude of interpatient pharmacokinetic variability for (131)I- and (111)In-Lym-1 suggested that an optimally safe and effective therapy can be best achieved when radionuclide dose is influenced by estimated radiation dose, if the latter is reproducible from institution to institution.