OBJECTIVE This study compares the effect of iodinated contrast agent on Hounsfield unit (HU)-based TG-186 dose calculation vs. delivered dose for high-dose-rate (HDR) iridium-192 brachytherapy using a phantom model. METHODS A reservoir filled with a diluted contrast agent was placed inside a water phantom. A single steel needle applicator was centrally positioned inside the reservoir. Computed tomography (CT) datasets of five different contrast agent dilutions (25 to 300 mg/ml iodine concentration) were acquired, and dose calculations were performed with TG-186 ACE dose calculation formalism of Oncentra®Brachy (Elekta). The dose was measured with a PinPoint® ionization chamber (PTW) inside the contrast agent. ACE calculated and measured data were compared. RESULTS For the different contrast agent dilutions, averaged Hounsfield units from 453 ±21 to 2623 ±221 were obtained. Electron densities derived from CT data were significantly higher than corresponding electron densities calculated from chemical compositions. Consequently, the measured dose was higher than corresponding HU-based calculated dose. Relative deviation ranged from 2.5% to 7% per 10 mm penetration depth, depending on contrast agent concentration. CONCLUSIONS The application of HU-based TG-186 dose formalisms in the presence of high-Z contrast agent bulks overestimates electron densities. Consequently, HU-based dose calculations result in a higher delivered dose than expected from the treatment planning system.