Abstract: The conventional neoadjuvant chemoradiotherapy (nCRT) combined with total mesorectal excision (TME) for locally advanced rectal cancer (LARC) yields a pathological complete response (pCR) rate of only 15%–20%, and radical surgery severely impairs patients’ quality of life. In recent years, the combination of immune checkpoint inhibitors (ICIs), represented by PD-1 monoclonal antibodies, with chemoradiotherapy has achieved breakthrough progress in proficient mismatch repair/microsatellite stable (pMMR/MSS) rectal cancer. Multiple prospective clinical studies, including our team’s POLARSTAR trial and NECTAR study, have confirmed that this combined strategy can increase the pCR rate to 30%–40%, and the 3-year disease-free survival (DFS) rate among patients achieving pCR can reach 100%. On this basis, how to accurately predict treatment response and safely implement organ preservation strategies has become a core issue in current research. This review discusses two dimensions: a multidimensional efficacy prediction system (including multiparametric MRI and radiomics, pathological TRG grading, NAR score, gut microbiota SPEED model, and circulating tumor DNA/minimal residual disease detection) and organ preservation strategies (watch-and-wait, local excision, and indications for TME). It also looks forward to the future direction of deep integration between precision prediction and organ preservation, aiming to promote the evolution of rectal cancer treatment toward a new paradigm that emphasizes both cure and function preservation.