To compare the results of axial length (AL) biometry in cataract eyes by three methods: immersion B-ultrasound (IB) biometry, immersion A-ultrasound (IA) biometry and optical low coherence reflectometry. In this prospective observational study of eyes with cataract AL measurements were performed using immersion ultrasound and optical low coherence reflectometry device. The results were evaluated using Bland-Altman analyses. The differences between both methods were assessed using the paired t-test, and its correlation was evaluated by Pearson coefficient. Eighty eyes of 80 patients (39 men and 41 women) for cataract surgery were included in the study. The values of AL could be got from all 80 eyes by IB and IA, the difference of AL measurements between IA and IB was of no statistical significance (P=0.97); the mean difference in AL measurements was -0.031 mm (P=0.26; 95%CI, -0.09 to 0.02); linear regression showed an excellent correlation (r=0.98, P<0.0001). Forty-five of eighty eyes with results of AL measurements, which can be obtained by three methods; the difference of AL measurements was of no statistical significance (IA vs IB, P=0.18; IA vs Lenstar, P=0.51; IB vs Lenstar, P=0.07); linear regression showed an excellent correlation (IA vs IB, r=0.99; IA vs Lenstar, r=0.96; IB vs Lenstar, r=0.96); Bland-Altman analysis also showed good agreement between the two methods [IA vs IB, 95% limits of agreement (LoA), -0.36 to 0.28 mm; IA vs Lenstar, 95% LoA, -0.65 to 0.69 mm; IB vs Lenstar, 95% LoA, -0.55 to 0.68 mm]. Measurements with the optical low coherence reflectometry correlated well with IB and IA. In the eyes with serious refractive medium opacity, the measurements of AL could not be achieved or existed deviations when using optical low coherence reflectometry device. Under such circumstances, we should choose IA or IB as the optimization method to obtain measurements, in order to get much more accurate results.