Background: To compare online adaptive radiation therapy (ART) to a clinically implemented plan selection strategy (PS) with respect to dose to the organs at risk (OAR) for rectal cancer. Methods: The first 20 patients treated with PS between May-September 2016 were included. This resulted in 10 short (SCRT) and 10 long (LCRT) course radiotherapy treatment schedules with a total of 300 Conebeam CT scans (CBCT). New dual arc VMAT plans were generated using auto-planning for both the online ART and PS strategy. For each fraction bowel bag, bladder and mesorectum were delineated on daily Conebeam CTs. The dose distribution planned was used to calculate daily DVHs. Coverage of the CTV was calculated, as defined by the dose received by 99% of the CTV volume (D99%). The volume of normal tissue irradiated with 95% of the prescribed fraction dose was calculated by calculating the volume receiving 95% of the prescribed fraction or more dose minus the volume of the CTV. For each fraction the difference between the plan selection and online adaptive strategy of each DVH parameter was calculated, as well as the average difference per patient. Results: Target coverage remained the same for online ART. The median volume of the normal tissue irradiated with 95% of the prescribed dose dropped from 642 cm3 (PS) to 237 cm3 (online-ART)(p < 0.001). Online ART reduced dose to the OARs for all tested dose levels for SCRT and LCRT (p < 0.001). For V15Gy of the bowel bag the median difference over all fractions of all patients was − 126 cm 3 in LCRT, while the average difference per patient ranged from − 206 cm 3 to − 40 cm 3. For SCRT the median difference was − 62 cm 3 , while the range of the average difference per patient was − 105 cm3 to − 51 cm 3. For V15Gy of the bladder the median difference over all fractions of all patients was 26% in LCRT, while the average difference per patient ranged from − 34 to 12%. For SCRT the median difference of V95% was − 8%, while the range of the average difference per patient was − 29 to 0%. Conclusions: Online ART for rectal cancer reduces dose the OARs significantly compared to a clinically implemented plan selection strategy, without compromising target coverage. Trial registration: Medical Research Involving Human Subjects Act (WMO) does not apply to this study and was retrospectively approved by the Medical Ethics review Committee of the Academic Medical Center (W19_357 # 19.