The main problem in total body irradiation (TBI) is obtaining a homogenous dose distribution inside the whole irradiated body and ensuring appropriate dose reduction in the lungs. The process of irradiation should be comfortable for the patients and repeatable despite the size and age diversity among patients. The aim of this paper was 1) to check accuracy of the applied dose algorithm and reliability of the measurement technique used in the lung region during TBI taken alternatively on a Cobalt-60 unit and on 15 MV linear accelerators, and 2) to check if the described methodology made it possible to obtain reproducibly of the lowered level of the dose to the lungs for a diverse group of patients. TBI was performed as a preparatory regiment in children and adults with disseminated malignancies undergoing bone marrow transplantation (a dose of 12.6 Gy in the midline/central beam axis). Two consecutive groups of patients were retrospectively included in the study: 15 irradiated with Cobalt-60 and 15 with 15 MV photons. The doses were evaluated for three sections passing through the middle of the lungs and at their upper and lower sides. Two types of detectors: semiconductor and thermoluminescent ones were used simultaneously. The measured doses were evaluated statistically to reveal agreement between readings of the two types of detectors and agreement between the measured doses and those previously calculated. The results of measurements exhibited a not Gaussian-type distribution (dissymmetry). The Wilcoxon-type test revealed compliance between the doses measured with thermoluminescent (TL) and semiconductor (SEM) detectors for all sections passing through the lungs (p>0.05), excluding the lung exit (middle and lower sides) with the Cobalt therapy. The t-Student test used to compare the measured doses with those previously calculated revealed agreement (p>0.05) between the measured doses and those calculated for all lung sections for the 15 MV photon therapy, while for Cobalt therapy such an agreement was at some points doubtful. The calculation algorithm and measurement techniques have proved to be correct, which was revealed by agreement between the doses measured and those calculated. The shielding of the lungs during both types of fields was effective and reproducible as indicated by agreement between the doses measured with the two types of detectors. Better agreement between the measured and calculated doses was found for 15 MV photons than for the Cobalt unit.