The aim of this study was to establish a fast and simple online dosimetric verification method for total body irradiation (TBI) using semiconductor diodes. Twenty one patients were treated with fractionated TBI (12 Gy, 2 Gy BID) using a modified standing technique between February 2010 and March 2012. The prescribed dose was administered to the patient's mid-plane at the level of the umbilicus. The lungs were shielded to allow only the absorption of an 8 Gy dose. To verify the dose administered during TBI, 12 semiconductor diodes were attached to the patient's skin to measure the entrance and exit doses at six anatomical reference points (umbilicus, lungs, neck, forehead, hand, and knee). An adapted version of the arithmetic method which uses a correction factor was used to calculate the mid-plane doses from the entrance and exit dose measurements. The mid-plane doses were then compared with the expected doses. The results of in vivo dosimetry showed that the mid-plane dose calculations conducted using the adapted arithmetic method were consistent with the expectations of the treatment plan. The average percentage dose differences and standard deviations were -0.5 +/- 5.3 % for the neck, -2.1 +/- 3.7 % for the umbilicus, 2.8 +/- 5.2 % for the forehead, and -5.9 +/- 11.1 % for the lungs. The extremity doses (hand and knee) were not calculated in the treatment planning program; however they were measured for documentary purposes. In vivo dose verification under TBI conditions was efficiently conducted through the use of semiconductor diodes. The measurement results were consistent with treatment plan dose calculations. The modified standing technique provides a practical method for TBI treatments. The observed heterogeneity is acceptable, the technique does not require additional shields or bolus material, the dosimetric verification is simple and fast, the set-up of the patient is easily reproducible, and the treatment time is within the acceptable limits for a non-dedicated treatment machine.