The use of an air-filled rectal balloon has been shown to decrease prostate motion during prostate radiotherapy. However, the perturbation of radiation dose near the air-tissue interfaces has raised clinical concerns of underdosing the prostate gland. The aim of this study was to investigate the dosimetric effects of an air-filled rectal balloon on the rectal wall/mucosa and prostate gland. Clinical rectal toxicity and dose-volume histogram (DVH) were also assessed to evaluate for any correlation. A film phantom was constructed to simulate the 4-cm diameter air cavity created by a rectal balloon. Kodak XV2 films were utilized to measure and compare dose distribution with and without air cavity. To study the effect in a typical clinical situation, the phantom was computed tomography (CT) scanned on a Siemens DR CT scanner for intensity-modulated radiation therapy (IMRT) treatment planning. A target object was drawn on the phantom CT images to simulate the treatment of prostate cancer. Because patients were treated in prone position, the air cavity was situated superiorly to the target. The treatment used a serial tomotherapy technique with the Multivane Intensity Modulating Collimator (MIMiC) in arc treatment mode. Rectal toxicity was assessed in 116 patients treated with IMRT to a mean dose of 76 Gy over 35 fractions (2.17-Gy fraction size). They were treated in the prone position, immobilized using a Vac-Loktrade mark bag and carrier-box system. Rectal balloon inflated with 100 cc of air was used for prostate gland immobilization during daily treatment. Rectal toxicity was assessed using modifications of the Radiation Therapy Oncology Group (RTOG) and late effects Normal Tissue Task Force (LENT) scales systems. DVH of the rectum was also evaluated. From film dosimetry, there was a dose reduction at the distal air-tissue interface as much as 60% compared with the same geometry without the air cavity for 15-MV photon beam and 2x2-cm field size. The dose beyond the interface recovered quickly and the dose reductions due to air cavity were 50%, 28%, 11%, and 1% at 2, 5, 10, and 15 mm, respectively, from the distal air-tissue interface. Evaluating the dose profiles of the more clinically relevant situation revealed the dose at air-tissue interface was approximately 15% lower in comparison to that without an air cavity. The dose built up rapidly so that at 1 and 2 mm, there was only an 8% and 5% differential, respectively. The dosimetric coverage at the depth of the posterior prostate wall was essentially equal with or without the air cavity. The median follow-up was 31.3 months. Rectal toxicity profile was very favorable: 81% (94/116) patients had no rectal complaint while 10.3% (12/116), 6.9% (8/116), and 1.7% (2/116) had grade 1, 2, and 3 toxicity, respectively. There was no grade 4 rectal toxicity. DVH analysis revealed that none of the patients had more than 25% of the rectum receiving 70 Gy or greater. Rectal balloon has rendered anterior rectal wall sparing by its dosimetric effects. In addition, it has reduced rectal volume, especially posterior and lateral rectal wall receiving high-dose radiation by rectal wall distension. Both factors may have contributed to decreased rectal toxicity achieved by IMRT despite dose escalation and higher than conventional fraction size. The findings have clinical significance for future very high-dose escalation trials whereby radiation proctitis is a major limiting factor.