The aim of this paper was to determine anthropometric parameters leading to the least uncertain estimate of heart size when connecting a computational phantom to an external beam radiation therapy (EBRT) patient. From computed tomography images, we segmented the heart and calculated its total volume (THV) in a population of 270 EBRT patients of both sexes, aged 0.7–83 years. Our data were fitted using logistic growth functions. The patient age, height, weight, body mass index and body surface area (BSA)were used as explanatory variables. For both genders, good fits were obtained with both weight (R2 = 0.89 formales and 0.83 for females) and BSA (R2=0.90 formales and 0.84 for females). These results demonstrate that, among anthropometric parameters, weight plays an important role in predicting THV. These findings should be taken into account when assigning a computational phantom to a patient.