Objective  To study and compare the effects of a 2.2 mm or 2.75 mm coaxial clear corneal incision （CCI） phacoemulsification on changes in corneal astigmatism. Methods  A prospective， randomized and controlled clinical study was conducted on 167 eyes of 134 cataract patients who were randomly divided into 2 groups： a 2.2 mm group （70 eyes of 57 patients） and a 2.75 mm group （97 eyes of 77 patients）. A superior incision was made when preoperative corneal astigmatism was <0.50 D， while an incision was made at the highest refractive power meridian when corneal astigmatism was ≥0.50 D. Uncorrected visual acuity （UCVA）， best corrected visual acuity （BCVA）， corneal astigmatism （CAD）， and corneal astigmatism axis （CAA）， were measured in both groups before the surgery and 1 week， 1 month and 3 months after surgery. A comparison and analysis between and within groups were performed on the above indicators at each time. Data were analyzed using an analysis of variance of the repeated measurement data， an independent samples t test， a group design Wilcoxon rank sum test， and an independent sample R×C contingency table χ² test. Results  Postoperative CAD decreased significantly in both groups， 0.73±0.43 D （preoperatively） changed to 0.49±0.36 D（3 months postoperatively） （P＜0.01） in the 2.2 mm group； 0.87±0.57 D （preoperatively） decreased to 0.53±0.38 D （3 months postoperatively） （P＜0.01） in the 2.75 mm group. The CAD of the 2.2 mm group briefly increased 1 week postoperatively （P＜0.05）， then decreased to preoperative levels （P＞0.05）. There was no significant change in the CAD of the 2.75 mm group 1 week postoperatively （P＞0.05）， then there was a significant reduction 1 month postoperatively （P＜0.05）. There was no significant difference in the changes in CAD amplitude between the groups. The overall distribution of CAA within groups showed no significant difference before and after surgery， and no significant difference between groups. Conclusion  By making CCI at the highest refractive power meridian， both the 2.2 mm and 2.75 mm CCI could partially correct the