Purpose:Large setup variability has been observed for prone breast patients due to rotation error. To reduce this uncertainty, we propose a novel implementation of isocentric prone breast method on Elekta linac with couch move assistant (CMA) and on‐line KV CBCT.Methods:Daily CBCT is used to evaluate the prone breast patient positioning uncertainty of proposed isocentric technique against our routine manual clinical setup. Clinical setup involves a manual AP shift from the ipsilateral torso tattoos (2PT) to the treatment iso directly underneath and SSD and flash check. While the proposed isocentric prone breast method features an additional contralateral leveling tattoo positioned at mid‐level of torso during CT simulation to determine correct patient obliqueness, and an automatic couch shift using Elekta CMA to correct both known distance from 2PT to the iso and daily setup uncertainty. Summary statistics were calculated for a cohort of prone breast patients from our clinic (n=5), and will be updated as more patients get administrated.Results:Small field of view partial CBCT acquisition is optimized to achieve clearance and minimize nominal imaging dose to 0.6cGy per scan to a 16cm phantom. Routine clinical setup uncertainty is 0.3±0.2cm, 0.7±0.4cm and 0.4±0.3cm in S/I, L/R and A/P respectively. Using the isocentric prone breast method, the setup uncertainty could be reduced to 0±0.3cm, 0.3±0.4cm and 0.1±0.3cm if the average couch shift in the 1st week of treatment is applied by CMA to the rest of treatment days. The uncertainty could be further reduced to 0.3±0.1cm, 0±0.1cm and 0±0.2cm with additional contralateral leveling tattoo. The improvement in precision is found to be statistically significant (p<0.05) using unpaired student t‐tests.Conclusion:Contralateral leveling tattoo is essential to determine correct patient obliqueness. More consistent and accurate isocentric prone breast patient positioning is achievable on Elekta linac with CMA and CBCT.