Purpose: To quantitatively validate and clinically implement a new algorithm, TMR10 for the LGK radiosurgery for brain cancer patients.Materials and Methods: Using actual patient MRI images, three individual plans were generated for 4, 8, and 16 mm cones by using TMR Classic. Plans with a single isocenter using same prescription, dose rate, grid size and shot location; were re‐computed using TMR10. Also, fifteen patients datasets that contained multiple isocenters (shots ranged from 1 to 65), who underwent GK radiosurgery using TMR Classic were retrospectively recalculated using TMR10. For these studies, the difference in total treatment time, treated volume under 50% and 90% isodose lines (50%, 90%IDV) were pairwise compared to TMR Classic plans. Also, for single isocenter studies, OAR points are defined 2cm away from the isocenter in three directions and point doses were calculated. Whereas qualitative assessment of 50% and 90%IDL were perform for the patient cases. Results: For the single isocenter studies, the largest difference was for the 8mm cone size in which total treatment time was increased by 3%, whereas 50% and 90%IDV were 8% and 27%, respectively. OAR doses were up to 12% higher when using TMR10. For the patient cases, with TMR10, total treatment time increased by 3.5%, whereas 50% and 90%IDV were 4% and 18%, respectively, in an average and maximum 90%IDV was up to 53%. Visual inspection of 50% and 90%IDL for both algorithms was similar, but, an effective prescription dose was decreased by nearly 3.5% when applying TMR10.Discussion: Dosimetric differences between the two algorithms were the characteristics of the more accurate Monte Carlo estimate of the new beam profiles, depth dose modeling, and output factors used by TMR10. TMR10 showed longer treatment time, higher OAR doses, and larger 50% and 90%IDV, but, 3.5% lower dose than the equivalent TMR classic.