Treatment with radioiodine is a standard procedure for patients with well-differentiated thyroid cancer, but the main approach to the therapy is still empiric, consisting of the administration of fixed activities. A predictive individualized dosimetric study may represent an important tool for physicians to determine the best activity to prescribe. The aim of this work is to compare red marrow and blood absorbed dose values obtained in the pre-treatment (PT) dosimetry phase with those obtained in the in-treatment (IT) dosimetry phase in order to estimate the predictive power of PT trial doses and to determine if they can be used as a decision-making tool to safely administer higher (131)I activity to potentially increase the efficacy of treatment. The PT and IT dosimetry for 50 patients has been evaluated using three different dosimetric approaches. In all three approaches blood and red marrow doses, are calculated as the sum of two components, the dose from (131)I activity in the blood and the dose from (131)I activity located in the remainder of the body (i.e. the blood and whole-body contributions to the total dose). PT and IT dose values to blood and red marrow appear to be well correlated irrespective of the dosimetric approach used. Linear regression analyses of PT and IT total doses, for blood and red marrow, and the whole-body contribution to these doses, showed consistent best fit slope and correlation coefficient values of approximately 0.9 and 0.6, respectively: analyses of the blood dose contribution to the total doses also yielded similar values for the best fit slope but with correlation coefficient values of approximately 0.4 reflecting the greater variance in these dose estimates. These findings suggest that pre-treatment red marrow dose assessments may represent an important tool to personalize metastatic thyroid cancer treatment, removing the constraints of a fixed activity approach and permitting potentially more effective higher (131)I activities to be safely used in-treatment.