Objective To investigate the factors that influence postoperative refractive error using the SRK/T and optimized Haigis formulas for eyes with high myopia, and to compare the accuracy of the two formulas. Methods This was a retrospective study. Thirty-eight patients (63 eyes) with axial lengths of ≥26 mm underwent cataract surgery from July 2011 to March 2015 in the Eye Hospital of Wenzhou Medical University. Axial length (AL), corneal curvature (K) and preoperative anterior chamber depth (AC) were measured by IOL-Master preoperatively. The SRK/T and optimized Haigis formulas were used to determine the intraocular lens (IOL) power and monofocal foldable IOLs were then implanted during surgery. Refractions were measured with a phoropter three months postoperatively. A paired t test was used to analyze the difference in refractive error measurements calculated by the two formulas. The difference in the proportion of hyperopia migration was analyzed by a McNemar test while the difference in the distribution of absolute error (AE) was analyzed by a Wilcoxon signed-rank test. The correlation of the difference in target refractions between the two formulas, as well as AE, AL, corneal curvature and preoperative anterior chamber depth were analyzed by a Pearson correlation analysis. Multiple linear regression was used to investigate the relative factors in refractive error based on the two formulas. Results Refractive errors calculated by the SRK/T and optimized Haigis formulas were -0.52±0.79 D and -0.67±0.79 D, respectively, with ratios of a hyperopic migration of 78% and 86%, respectively. The difference in target refraction between the two formulas was negatively correlated with AL (r=-0.27, P<0.05), and positively correlated with K (r=0.73, P<0.01). Both AEs were positively correlated with axial length (SRK/T: r=0.43, P<0.01; optimized Haigis: r=0.31, P<0.05). Anterior chamber depth was grouped into AC≤3.5 mm and AC>3.5 mm. In the group with AC≤3.5 mm, linear regression formulas for refractive errors calculated with the SRK/T and optimized Haigis formula were ■=3.74-0.15AL (R2=0.27, F=3.88, P<0.05) and ■=12.03-0.19K-0.13AL (R2=0.27, F=3.73, P<0.05). In the group with AC>3.5 mm, linear regression formulas for refractive errors with the SRK/T and optimized Haigis formula were ■=1.40AC-0.37AL (R2=0.62, F=13.40, P<0.01) and ■=14.02+1.76AC-0.25K-0.35AL (R2=0.62, F=13.59, P<0.01). Conclusion For eyes with a long axial length, the optimized Haigis formula is better than the SRK/T formula because postoperative refractive error exhibits a more obvious hyperopic shift using the SRK/T formula. With regard to a steep K value, refractive error increases as K augments using the optimized Haigis formula. Thus the SRK/T formula is the better choice under this condition. When preoperative anterior chamber depth is more than 3.5 mm, the target myopic refraction can be reduced with either formula.