Background Cataract is the main cause of global blindness,and surgery is the main method for the treatment of cataract,so precise preoperative parameters have an important significance for improving postoperative visual quality.Objective This study was to analyze the difference of lens thickness (LT) measured by immersion A-scan ultrasound versus optical biometry and consistency of these two methods in measuring the LT for different types of cataract eyes,and to investigate whether A-scan ultrasound for the LT measurement needs to correct the ultrasound speed parameter.Methods A reliability evaluation of diagnosis test was designed.The LT was measured on 143 cataract eyes using immersion A-scan ultrasound and Lenstar optical biometry in Tianjin Eye Hospital from September 2013 to May 2014.The eyes were assigned to nuclear cataract group (60 eyes of 50 patients),posterior subcapsular cataract group (40 eyes of 40 patients) and cortical cataract group (43 eyes of 43 patients) based on the types of cataract,and the hardness of nucleus was grade II in 20 eyes of 14 patients,grade Ⅲ in 20 eyes of 17 patients and grade Ⅳ in 20 eyes of 19 patients.The outcomes measured by immersion A-scan ultrasound versus Lenstar optical biometry were compared with paired t test,and the consistency between the two devices was analyzed by Bland-Altman agreement plot.This study complied with Helsinki declaration,and written informed consent was obtained from each patient prior to any examination.Results The mean LT value was (4.85± 0.39) mm from the immersion A-scan ultrasound and (4.74 ±0.37)mm from the Lenstar optical biometry,with a significant difference between the two devices (t =3.020,P =0.004).No significant differences were found in the LT values of the posterior subcapsular cataractous eyes and cortical cataractous eyes between the two devices (t=1.015,P =0.316;t =1.275,P =0.209).The LT values by immersion A-scan ultrasound were significantly higher in grade Ⅱ,Ⅲ and Ⅳ nuclear cataractous eyes than those by Lenstar optical biometry (t=2.175,2.378,2.383,all at P＜0.05).The outcome showed a good consistency in the posterior subcapsular cataractous eyes between the two measuring methods,with the 95％ limitation of agreement (LoA) from-0.21 to 0.18 mm.However,there were poor consistencies between the two measuring methods in the nuclear and cortical cataractous eyes,with the 95％ LoA from-0.64 to 0.43 mm and from-0.50 to 0.41 mm,respectively.Poor consistencies also were seen in grade Ⅱ,Ⅲ,Ⅳ nuclear cataractous eyes between the two measuring methods,and the 95％ LoA were-0.31 to 0.22 mm,-0.32 to 0.24mm and-1.09 to 0.62 mm,respectively.Conclusions There are differences in measuring LT of different types of cataractous eyes between immersion A-scan ultrasound and Lenstar optical biometry.The measuring values are higher by immersion A-scan ultrasound than those by Lenstar optical biometry,suggesting that the measurement of immersion A-scan ultrasound for LT of different types and nuclear hardness of cataractous eyes should correct the ultrasound speed parameters.