PURPOSE: Comparing IOL power calculation formulas in long eyes (AL≥26.00 mm) to find the best axial length (AL) adjustment/IOL power calculation formula combination. DESIGN: Retrospective, comparative, case-series. PARTICIPANTS: Patients with long eyes that underwent cataract surgery. METHODS: five-hundred-fifty-four eyes of 554 patients were examined before and after standard phacoemulsification without complications. Eyes were subdivided in 3 groups according to AL: 26.00≤AL<28.00 mm, 28.00≤AL<30.00 mm, AL≥30.00 mm. Eight formulas that do not require anterior chamber depth (ACD) were evaluated: Barrett Universal II (BUII), Emmetropia Verifying Optical (EVO) 2.0, Ladas Super Formula (LSF), Hoffer Q, Holladay 1, SRKT, T2 and T2.2. The lens constant of ULIB database and IOLCon database were used. Each formula was analyzed by using uncorrected AL (ALu) and following AL adjustments: Wang-Koch 1 (wk1), wk2, wk polinomial (wk-pol), estimated Cooke modified axial length (CMALe) and ALc correcting factor. MAIN OUTCOME MEASURES: Mean absolute error (MAE), median absolute error (MedAE) and percentage of eyes within ±0.50 and ± 1.00 diopters (D) of prediction error. RESULTS: T2-ALu gave best outcome when 26.00 mm ≤ AL<28.00 mm. LSF-ALu, BUII-ALu, EVO 2.0-ALu, Holladay 1-wk-pol and T2.2-CMALe represented valid alternatives. EVO 2.0-ALc gave best outcomes when 28.00 mm ≤ AL<30.00 mm. Other thick-lens or hybrid artificial-intelligence-vergence based formulas (BUII-ALu, LSF-CMALe) and Holladay 1-wk2 demonstrated greater reliability compared to thin lens-based formulas. EVO 2.0-CMALe gave best outcomes when AL≥30.00 mm. Holladay 1-wk-pol e T2.2-wk1 represented valid alternatives (all p < 0.050). LSF could fail in 50 % of cases without ACD when AL≥30.00 mm. CONCLUSIONS: Choosing the best AL adjustment/IOL power calculation formula combination for each AL subrange, can improve refractive outcomes in patients with long eyes that undergo cataract surgery.