To assess the safety and effectiveness of subnuclear phacoemulsification (SNP), a new technique for reducing corneal injury in nuclear cataract surgery. This randomized controlled trial (RCT) was designed in March 2020 and carried it out from April 1 to September 30, 2020, including a 3 months' follow-up. We recruited 256 age-related hard nucleus cataract patients and randomly assigned them to two groups: the experimental group receiving SNP, and the control group receiving conventional phacoemulsification (CP). A single surgeon performed all the surgeries. We compared the two groups on the cumulative dissipated energy (CDE), phacoemulsification ultrasound time (UST), and complications for safety, as well as at multiple postsurgery follow-up timepoints on three major outcomes for effectiveness: visual acuity, central corneal thickness, and central corneal endothelial cell density. The two groups were well-matched in terms of demographics, nuclear density, and safety measures (ultrasound energy, phacoemulsification time, and complications). For effectiveness after surgery, compared to the CP group, the SNP group had better visual acuity and thinner central cornea postsurgery within 1 week and had higher central corneal endothelial cell density at the 1- and 3-month follow-up. Compared to CP, SNP is more effective for reducing corneal injury in cataract surgery. The widespread application of this technology will greatly improve the safety of cataract surgery, especially hard cataract surgery. Chinese Clinical Trial Registry: ChiCTR2000031114.